Skip to content

Instantly share code, notes, and snippets.

@thesps

thesps/plot_scan.py

Created April 19, 2021 10:51
Show Gist options
  • Save thesps/8a6dff86ec941e2515cdbd52cc337711 to your computer and use it in GitHub Desktop.
Save thesps/8a6dff86ec941e2515cdbd52cc337711 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
plot_scan.py
import hls4ml
from optparse import OptionParser
import pandas
import matplotlib.pyplot as plt
from matplotlib.ticker import MaxNLocator
from sklearn.metrics import roc_curve, auc, accuracy_score
from sklearn.preprocessing import OneHotEncoder
import numpy as np
import tensorflow as tf
from qkeras.utils import _add_supported_quantized_objects
from train import get_features, default_options_and_config
import yaml
hls4ml.model.optimizer.OutputRoundingSaturationMode.layers = ['Activation']
hls4ml.model.optimizer.OutputRoundingSaturationMode.rounding_mode = 'AP_RND'
hls4ml.model.optimizer.OutputRoundingSaturationMode.saturation_mode = 'AP_SAT'
avail = {'dsp' : 6840, 'lut' : 1182240, 'ff' : 2364480}
def get_Xyl():
'''Load the default dataset and return (X_test, y_test, labels) tuple'''
opts, cfg = default_options_and_config()
X_train_val, X_test, y_train_val, y_test, labels = get_features(opts, cfg)
ohe = OneHotEncoder().fit(np.arange(0,5,1).reshape(-1,1))
y_test = ohe.inverse_transform(y_test)
Xyl = (X_test, y_test, labels)
return Xyl
def model_performance_summary(model, hls_model, Xyl, report=False):
'''Get the performance summary (model and synthesis) for a model and corresponding hls_model
args:
model: a (Q)Keras model
hls_model: the corresponding hls4ml HLSModel object
Xyl: a tuple of (X_test, y_test, labels) e.g. returned by get_Xyl
report: boolean, whether or not to load the HLS synthesis reports
'''
X_test, y_test, labels = Xyl
y_predict = model.predict(X_test)
y_predict_hls4ml = hls_model.predict(X_test)
data = {}
one = np.ones_like(y_test.shape[0])
zero = np.zeros_like(y_test.shape[0])
for j in range(len(labels)):
label = labels[j].replace('j_','')
y_class = (y_test == j).flatten().reshape(y_test.shape[0])
y_binary = np.where(y_class, one, zero)
fpr, tpr, thresh = roc_curve(y_binary, y_predict[:,j])
data['auc_qkeras_' + label] = auc(fpr, tpr)
fpr, tpr, thresh = roc_curve(y_binary, y_predict_hls4ml[:,j])
data['auc_hls4ml_' + label] = auc(fpr, tpr)
data['accuracy_qkeras'] = accuracy_score(y_test, np.argmax(y_predict,axis=1).reshape(-1,1))
data['accuracy_hls4ml'] = accuracy_score(y_test, np.argmax(y_predict_hls4ml,axis=1).reshape(-1,1))
if report:
# Get the resources from the logic synthesis report
report = open('{}/vivado_synth.rpt'.format(hls_model.config.get_output_dir()))
lines = np.array(report.readlines())
data['lut'] = int(lines[np.array(['CLB LUTs*' in line for line in lines])][0].split('|')[2])
data['ff'] = int(lines[np.array(['CLB Registers' in line for line in lines])][0].split('|')[2])
data['bram'] = float(lines[np.array(['Block RAM Tile' in line for line in lines])][0].split('|')[2])
data['dsp'] = int(lines[np.array(['DSPs' in line for line in lines])][0].split('|')[2])
report.close()
# Get the latency from the Vivado HLS report
report = open('{}/myproject_prj/solution1/syn/report/myproject_csynth.rpt'.format(hls_model.config.get_output_dir()))
lines = np.array(report.readlines())
lat_line = lines[np.argwhere(np.array(['Latency (cycles)' in line for line in lines])).flatten()[0] + 3]
data['latency_clks'] = int(lat_line.split('|')[2])
data['latency_ns'] = float(lat_line.split('|')[4].replace('ns',''))
return data
def extract_data_scan(scan_dir, Xyl):
co = {}
_add_supported_quantized_objects(co)
data = {'w':[], 'accuracy_qkeras':[], 'accuracy_hls4ml' : [], 'dsp':[], 'lut':[], 'ff':[],
'bram':[], 'latency_clks':[], 'latency_ns':[]}
for label in ['q','g','t','w','z']: data['auc_qkeras_' + label] = []
for label in ['q','g','t','w','z']: data['auc_hls4ml_' + label] = []
for i in range(3,17):
model = tf.keras.models.load_model('{}/model_{}/KERAS_check_best_model.h5'.format(scan_dir, i), custom_objects=co)
hls4ml_cfg = hls4ml.utils.config_from_keras_model(model, granularity='name')
hls4ml_cfg['LayerName']['softmax']['exp_table_t'] = 'ap_fixed<18,9>'
hls4ml_cfg['LayerName']['softmax']['inv_table_t'] = 'ap_fixed<18,4>'
hls_model = hls4ml.converters.convert_from_keras_model(model, output_dir='{}/model_{}/hls4ml_prj'.format(scan_dir, i),
fpga_part='xcvu9p-flgb2104-2L-e', hls_config=hls4ml_cfg)
hls_model.compile()
datai = model_performance_summary(model, hls_model, Xyl, report=True)
for key in datai.keys():
data[key].append(datai[key])
data = pandas.DataFrame(data)
return data
def extract_data_baseline_unpruned(Xyl):
model = tf.keras.models.load_model('baseline/KERAS_check_best_model.h5')
hls4ml_cfg = hls4ml.utils.config_from_keras_model(model, granularity='Model')
data = {'w':[], 'accuracy_qkeras':[], 'accuracy_hls4ml' : [], 'dsp':[], 'lut':[], 'ff':[],
'bram':[], 'latency_clks':[], 'latency_ns':[]}
for label in ['q','g','t','w','z']: data['auc_qkeras_' + label] = []
for label in ['q','g','t','w','z']: data['auc_hls4ml_' + label] = []
for i in range(3,17):
hls4ml_cfg['Model']['Precision'] = 'ap_fixed<{},6>'.format(i)
hls_model = hls4ml.converters.convert_from_keras_model(model, output_dir='baseline/hls4ml_prj_{}'.format(i),
fpga_part='xcvu9p-flgb2104-2L-e', hls_config=hls4ml_cfg)
hls_model.compile()
datai = model_performance_summary(model, hls_model, Xyl, report=False)
for key in datai.keys():
data[key].append(datai[key])
return data
def extract_data_baseline(Xyl):
model = tf.keras.models.load_model('baseline/KERAS_pruned_model.h5')
hls4ml_cfg = open('baseline/hls4ml_cfg.yml')
hls4ml_cfg = yaml.load(hls4ml_cfg, Loader=yaml.SafeLoader)
hls_model = hls4ml.converters.convert_from_keras_model(model, output_dir='baseline/hls4ml_prj_eval',
fpga_part='xcvu9p-flgb2104-2L-e', hls_config=hls4ml_cfg)
hls_model.compile()
data = model_performance_summary(model, hls_model, Xyl, report=False)
return data
def make_plots(data):
f0 = plt.figure()
plt.plot(data['w'], data['dsp'] * 10, label=r'DSP \times 10')
plt.plot(data['w'], data['lut'], label=r'LUT')
plt.plot(data['w'], data['ff'], label=r'FF')
plt.legend()
plt.xlabel('Bitwidth')
plt.ylabel('Resource Usage')
f1 = plt.figure()
plt.plot(data['w'], data['dsp'] / avail['dsp'], label=r'DSP')
plt.plot(data['w'], data['lut'] / avail['lut'], label=r'LUT')
plt.plot(data['w'], data['ff'] / avail['ff'], label=r'FF')
plt.legend()
plt.xlabel('Bitwidth')
plt.ylabel('Resource Usage (% VU9P)')
f2 = plt.figure()
plt.plot(data['w'], data['accuracy_qkeras'], label='QKeras')
plt.plot(data['w'], data['accuracy_hls4ml'], label='hls4ml')
plt.xlabel('Bitwidth')
plt.ylabel('Accuracy')
return f0, f1, f2
def do_plots(percent_resources=False, log=False):
import pandas; import plt4hls4ml; plt4hls4ml.init()
data = pandas.read_csv('scan_models/summary.csv')
data0 = pandas.read_csv('baseline/summary_eps1_rising_2.csv')
data1 = pandas.read_csv('baseline_junior/summary_optimized.csv')
baseline = pandas.read_csv('baseline/summary_unpruned_scan.csv')
baseline0 = pandas.read_csv('baseline/summary_full_14_6.csv')
baseline1 = pandas.read_csv('baseline/summary_pruned_14_6.csv')
symbols = ['P', 'o', 'X', 'D']
d = [baseline0, baseline1, data0, data1]
names = ['B', 'BP', 'BH', 'QO']
norm_lut = 100. / avail['lut'] if percent_resources else 1
norm_ff = 100. / avail['ff'] if percent_resources else 1
norm_dsp = 100. / avail['dsp'] if percent_resources else 10
width = [3,1]
f = plt.figure(figsize=(3,3), constrained_layout=True)
gs = f.add_gridspec(ncols=2, nrows=1, width_ratios=width)
ax0 = f.add_subplot(gs[0,0])
plt.plot(data['w'], data['lut'] * norm_lut, label=r'LUT')
plt.plot(data['w'], data['ff'] * norm_ff, label=r'FF')
plt.plot(data['w'], data['dsp'] * norm_dsp, label=r'DSP' if percent_resources else r'DSP $\times 10$' )
plt.gca().set_prop_cycle(None)
#plt.plot(data[data['w'] == 6]['w'], data[data['w'] == 6]['lut'] * norm_lut, 'x')
#plt.plot(data[data['w'] == 6]['w'], data[data['w'] == 6]['ff'] * norm_ff, 'x')
#plt.plot(data[data['w'] == 6]['w'], data[data['w'] == 6]['dsp'] * norm_dsp, 'x')
#plt.gca().ticklabel_format(axis='y', style='scientific', scilimits=(0,0))
plt.legend()
plt.xlabel('Bitwidth')
plt.ylabel('Resource Usage (%)')
plt.gca().get_xaxis().set_major_locator(MaxNLocator(integer=True))
ax1 = f.add_subplot(gs[0,1], sharey=ax0)
x = 0
plt.gca().set_prop_cycle(None)
for i, di in enumerate(d):
s = symbols[i]
plt.plot(x, di['lut'] * norm_lut, s)
plt.plot(x, di['ff'] * norm_ff, s)
plt.plot(x, di['dsp'] * norm_dsp, s)
x += 1
plt.gca().set_prop_cycle(None)
ax1.set_xlim((-0.5, len(d) - 0.5))
ax1.get_yaxis().set_visible(False)
plt.xticks(list(range(len(names))), names, fontsize=6)
if log:
plt.semilogy()
plt.savefig('resources.pdf')
f1 = plt.figure(figsize=(3,3), constrained_layout=True)
gs = f1.add_gridspec(ncols=2, nrows=1, width_ratios=width)
ax0 = f1.add_subplot(gs[0,0])
plt.plot(data['w'], data['accuracy_qkeras'] / baseline['accuracy_qkeras'].iloc[0], label='QKeras CPU')
plt.plot(data['w'], data['accuracy_hls4ml'] / baseline['accuracy_qkeras'].iloc[0], label='QKeras FPGA')
plt.plot(baseline['w'], baseline['accuracy_hls4ml'] / baseline['accuracy_qkeras'], '--', label='Post-train quant.')
plt.gca().set_prop_cycle(None)
#plt.plot(data[data['w'] == 6]['w'], data[data['w'] == 6]['accuracy_qkeras'] / baseline['accuracy_qkeras'].iloc[0], 'x')
#plt.plot(data[data['w'] == 6]['w'], data[data['w'] == 6]['accuracy_hls4ml'] / baseline['accuracy_qkeras'].iloc[0], 'x')
plt.xlabel('Bitwidth')
plt.ylabel('Ratio Model Accuracy / Baseline Accuracy')
plt.ylim((0.9, 1.05))
plt.legend(loc='upper left')
ax1 = f1.add_subplot(gs[0,1], sharey=ax0)
plt.gca().set_prop_cycle(None)
x = 0
c1 = plt4hls4ml.colors[1]
c2 = plt4hls4ml.colors[2]
colors = [c2, c2, c2, c1]
for i, di in enumerate(d):
s = symbols[i]
plt.plot(x, di['accuracy_hls4ml'] / baseline['accuracy_qkeras'].iloc[0], s, color=colors[i])
#plt.gca().set_prop_cycle(None)
x += 1
ax1.set_xlim((-0.5, len(d) - 0.5))
ax1.get_yaxis().set_visible(False)
plt.xticks(list(range(len(names))), names, fontsize=6)
plt.savefig('accuracy.pdf')
f1 = plt.figure(figsize=(3,3), constrained_layout=True)
gs = f1.add_gridspec(ncols=2, nrows=1, width_ratios=width)
ax0 = f1.add_subplot(gs[0,0])
for label in ['g', 'q', 'w', 'z', 't']:
y = data['auc_hls4ml_{}'.format(label)] / baseline['auc_qkeras_{}'.format(label)]
plt.plot(data['w'], y, label=label)
plt.legend()
plt.gca().set_prop_cycle(None)
for label in ['g', 'q', 'w', 'z', 't']:
y = baseline['auc_hls4ml_{}'.format(label)] / baseline['auc_qkeras_{}'.format(label)]
plt.plot(baseline['w'], y, '--')
plt.xlabel('Bitwidth')
plt.ylabel('Ratio AUC QKeras+hls4ml / Keras float')
ax1 = f1.add_subplot(gs[0,1], sharey=ax0)
plt.gca().set_prop_cycle(None)
for label in ['g', 'q', 'w', 'z', 't']:
y = data0['auc_hls4ml_{}'.format(label)] / baseline['auc_qkeras_{}'.format(label)].iloc[0]
plt.plot(0, y, 'P')
plt.gca().set_prop_cycle(None)
for label in ['g', 'q', 'w', 'z', 't']:
y = data1['auc_hls4ml_{}'.format(label)] / baseline['auc_qkeras_{}'.format(label)].iloc[0]
plt.plot(1, y, 'o')
for label in ['g', 'q', 'w', 'z', 't']:
y = data1['auc_hls4ml_{}'.format(label)] / baseline['auc_qkeras_{}'.format(label)].iloc[0]
plt.plot(1, y, 'o')
ax1.set_xlim((-0.5, 1.5))
ax1.get_yaxis().set_visible(False)
plt.xticks([0, 1], ['A', 'B'])
plt.savefig('auc.pdf')
if __name__ == "__main__":
parser = OptionParser()
parser.add_option('-d', '--dir', action='store', type='string', dest='directory', default='scan_models')
parser.add_option('-o', '--outfile', action='store', type='string', dest='outfile', default='summary.csv')
parser.add_option('-i', '--infile', action='store', type='string', dest='infile', default=None)
(options,args) = parser.parse_args()
outfile = options.directory + options.outfile
if options.infile is None:
Xyl = get_Xyl()
data = extract_data_scan(options.directory, Xyl)
data.to_csv(outfile)
else:
data = pandas.read_csv(options.pandasFile)
make_plots(data)
Display the source blob
Display the rendered blob
Raw
scan_cnn.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Display the source blob
Display the rendered blob
Raw
scan_plots.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import tensorflow as tf\n",
"import numpy as np\n",
"import hls4ml\n",
"seed = 0\n",
"np.random.seed(seed)\n",
"import tensorflow as tf\n",
"tf.random.set_seed(seed)\n",
"import os\n",
"os.environ['PATH'] = '/opt/Xilinx/Vivado/2019.2/bin:' + os.environ['PATH']\n",
"from joblib import Parallel, delayed\n",
"import sys\n",
"import pandas\n",
"import plt4hls4ml\n",
"plt4hls4ml.init()\n",
"from cycler import cycler\n",
"import matplotlib as mpl\n",
"import matplotlib.pyplot as plt\n",
"cols = ['#a6cee3','#1f78b4','#b2df8a','#33a02c','#fb9a99','#e31a1c','#fdbf6f','#ff7f00','#cab2d6','#6a3d9a','#ffff99']\n",
"mpl.rcParams['axes.prop_cycle'] = cycler(color=cols)\n",
"from matplotlib.patches import Patch\n",
"from matplotlib.lines import Line2D"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Scan 1"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Calculate some image sizes\n",
"area = np.linspace(100, 256*256, 20)\n",
"side = np.round(np.sqrt(area)).astype('int')\n",
"area = np.square(side).astype('int')\n",
"channels = 3\n",
"n_filters = np.linspace(1, 20, 20).astype('int')\n",
"kernel_size = np.array([3, 5, 7])\n",
"N = int(len(side) * len(kernel_size) * len(n_filters))\n",
"sides_kernels_filters = np.array(np.meshgrid(side, kernel_size, n_filters)).T.reshape(N, 3)\n",
"scan_dir = 'scan_single_c2d'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Scan 2"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"# Calculate some image sizes\n",
"area = np.linspace(100, 256*256, 20)\n",
"side = np.round(np.sqrt(area)).astype('int')\n",
"area = np.square(side).astype('int')\n",
"channels = 3\n",
"n_filters = np.array([1])\n",
"n_filters = np.append(n_filters, np.arange(0, 36, 4).astype('int')[1:])\n",
"n_filters = np.append(n_filters, np.array([48, 64]))\n",
"kernel_size = np.array([3, 5, 7])\n",
"N = int(len(side) * len(kernel_size) * len(n_filters))\n",
"sides_kernels_filters = np.array(np.meshgrid(side, kernel_size, n_filters)).T.reshape(N, 3)\n",
"scan_dir = 'scan_single_c2d_2'"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [],
"source": [
"def report(proj_dir):\n",
" data = {}\n",
" data['error'] = False\n",
" \n",
" # Check for C Synthesis errors\n",
" if os.path.exists('{}/vivado_hls.log'.format(proj_dir)):\n",
" report = open('{}/vivado_hls.log'.format(proj_dir))\n",
" lines = np.array(report.readlines())\n",
" error = bool(np.sum(np.array(['ERROR' in line for line in lines])))\n",
" if error:\n",
" data['latency_us'] = 0\n",
" data['latency_clks'] = 0\n",
" data['lut'] = 0\n",
" data['ff'] = 0\n",
" data['bram'] = 0\n",
" data['dsp'] = 0\n",
" data['error'] = True\n",
" return data\n",
" \n",
" # Get the latency from the Vivado HLS report\n",
" if os.path.exists('{}/myproject_csynth.rpt'.format(proj_dir)):\n",
" report = open('{}/myproject_csynth.rpt'.format(proj_dir))\n",
" lines = np.array(report.readlines())\n",
" lat_line = lines[np.argwhere(np.array(['Latency (cycles)' in line for line in lines])).flatten()[0] + 3]\n",
" data['latency_clks'] = int(lat_line.split('|')[2])\n",
" lat = lat_line.split('|')[4]\n",
" if 'ns' in lat:\n",
" # put in microseconds\n",
" lat = float(lat.replace('ns','')) * 10**-3\n",
" elif 'us' in lat:\n",
" lat = float(lat.replace('us',''))\n",
" elif 'ms' in lat:\n",
" lat = float(lat.replace('ms','')) * 10**3\n",
" else:\n",
" print(\"Unsupported latency suffix: {}\".format(lat))\n",
" data['latency_us'] = lat\n",
" else:\n",
" return None\n",
" \n",
" # Get the resources from the logic synthesis report\n",
" if os.path.exists('{}/vivado_synth.rpt'.format(proj_dir)):\n",
" report = open('{}/vivado_synth.rpt'.format(proj_dir))\n",
" lines = np.array(report.readlines())\n",
" data['lut'] = int(lines[np.array(['CLB LUTs*' in line for line in lines])][0].split('|')[2]) \n",
" data['ff'] = int(lines[np.array(['CLB Registers' in line for line in lines])][0].split('|')[2])\n",
" data['bram'] = float(lines[np.array(['Block RAM Tile' in line for line in lines])][0].split('|')[2]) \n",
" data['dsp'] = int(lines[np.array(['DSPs' in line for line in lines])][0].split('|')[2])\n",
" report.close()\n",
" else:\n",
" return None\n",
" \n",
" return data"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [],
"source": [
"def extract_data_scan(scan_dir, rang):\n",
"\n",
" data = {'im_height':[], 'im_width':[], 'n_filt':[], 'filt_height':[], 'filt_width':[],\n",
" 'dsp':[], 'lut':[], 'ff':[], 'bram':[], 'latency_clks':[], 'latency_us':[], 'error':[]}\n",
"\n",
" for i in rang:\n",
" datai = report(scan_dir + '/project_' + str(i))\n",
" if datai is not None:\n",
" datai['im_height'] = sides_kernels_filters[i][0]\n",
" datai['im_width'] = sides_kernels_filters[i][0]\n",
" datai['filt_height'] = sides_kernels_filters[i][1]\n",
" datai['filt_width'] = sides_kernels_filters[i][1]\n",
" datai['n_filt'] = sides_kernels_filters[i][2]\n",
" for key in datai.keys():\n",
" data[key].append(datai[key])\n",
" \n",
"\n",
" data = pandas.DataFrame(data)\n",
" return data"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [],
"source": [
"data = extract_data_scan(scan_dir, range(len(sides_kernels_filters)))"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [],
"source": [
"data.to_csv('{}/summary.csv'.format(scan_dir))"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {},
"outputs": [],
"source": [
"import matplotlib.pyplot as plt\n",
"%matplotlib inline"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [],
"source": [
"#d = data[data['n_filt'] % 2 == 0]\n",
"d = data[data['error'] == False]"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {},
"outputs": [],
"source": [
"linestyles = ['-', ':', '--', '-.']\n",
"def do_a_plot(y_key, y_label):\n",
" plt.figure(figsize=(9,9))\n",
" for ifh, fh in enumerate(np.unique(np.array(d.filt_height))):\n",
" for inf, nf in enumerate(np.unique(np.array(d.n_filt))):\n",
" di = d[d.filt_height == fh][d.n_filt == nf]\n",
" a = di['im_height'] * di['im_width']\n",
" plt.plot(a, di[y_key], '.', linestyle=linestyles[ifh],\n",
" color=cols[inf%len(cols)])#, label='kernel:{}x{}, n_filt:{}'.format(fh, fh, nf))\n",
" handles = []\n",
" for ifh, fh in enumerate(np.unique(np.array(d.filt_height))):\n",
" handles.append(Line2D([0], [0], color='k', marker='', linestyle=linestyles[ifh], label='kernel:{}x{}'.format(fh, fh)))\n",
" for inf, nf in enumerate(np.unique(np.array(d.n_filt))):\n",
" handles.append(Line2D([0], [0], color=cols[inf%len(cols)], marker='.', label='n_filt:{}'.format(nf)))\n",
" plt.legend(handles=handles, loc='upper right')\n",
" plt.xlabel('H x W')\n",
" plt.ylabel(y_label)\n",
" plt.savefig('{}/plots/HxWvs_{}.pdf'.format(scan_dir, y_key))\n",
" plt.savefig('{}/plots/HxWvs_{}.png'.format(scan_dir, y_key))"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 648x648 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"do_a_plot('dsp', 'DSPs')"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 648x648 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"do_a_plot('lut', 'LUTs')"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 648x648 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"do_a_plot('latency_us', 'Latency (μs)')"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {},
"outputs": [],
"source": [
"x = data[data.im_height == 10].n_filt\n",
"y = data[data.im_height == 10].filt_width\n",
"z = np.array(data[data.im_height == 10].error.apply(lambda x : 'red' if x else 'green'))"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 288x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(4,4))\n",
"for i in range(len(z)):\n",
" plt.plot(x.iloc[i], y.iloc[i], '.', color=z[i])\n",
" plt.text(x.iloc[i], y.iloc[i] + (-1)**(i%2) * 0.3, str(x.iloc[i] * 3 * y.iloc[i]**2),\n",
" horizontalalignment='center', verticalalignment='center', fontsize=7)\n",
"plt.xlabel('n_filters')\n",
"plt.ylabel('filt_width')\n",
"plt.ylim((2, 8))\n",
"plt.legend(handles=[Line2D([0], [0], color='g', marker='.', label='Success'),\n",
" Line2D([0], [0], color='r', marker='.', label='Fail')], ncol=2, loc='upper center')\n",
"plt.savefig('{}/plots/limit.pdf'.format(scan_dir))\n",
"plt.savefig('{}/plots/limit.png'.format(scan_dir))"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:1: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"Entry point for launching an IPython kernel.\n"
]
}
],
"source": [
"d = data[data.im_height.isin(side[np.array(range(int(len(side) / 4))) * 4])][data.error == False]"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [],
"source": [
"def do_a_plot_2(y_key, y_label):\n",
" plt.figure(figsize=(9,9))\n",
" for iw, w in enumerate(np.unique(np.array(d.im_width))):\n",
" for ifh, fh in enumerate(np.unique(np.array(d.filt_height))):\n",
" di = d[d.im_width == w][d.filt_height == fh]\n",
" x = di.n_filt\n",
" plt.plot(x, di[y_key], '.', linestyle=linestyles[ifh],\n",
" label='image:{}x{}, filter:{}x{}'.format(w, w, fh, fh), color=cols[iw%len(cols)])\n",
" handles = [] \n",
" \n",
" for iw, w in enumerate(np.unique(np.array(d.im_width))):\n",
" handles.append(Line2D([0], [0], color=cols[iw%len(cols)], marker='.', label='image:{}x{}'.format(w, w)))\n",
"\n",
" for ifh, fh in enumerate(np.unique(np.array(d.filt_height))):\n",
" handles.append(Line2D([0], [0], color='k', marker='', linestyle=linestyles[ifh], label='filter:{}x{}'.format(fh, fh)))\n",
"\n",
" plt.legend(handles=handles, loc='upper right')\n",
" plt.xlabel('n_filt')\n",
" plt.ylabel(y_label)\n",
" plt.savefig('{}/plots/n_filtvs_{}.pdf'.format(scan_dir, y_key))\n",
" plt.savefig('{}/plots/n_filtvs_{}.png'.format(scan_dir, y_key))"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 648x648 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"do_a_plot_2('dsp', 'DSPs')"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 648x648 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"do_a_plot_2('lut', 'LUTs')"
]
},
{
"cell_type": "code",
"execution_count": 38,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/ipykernel_launcher.py:5: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
" \"\"\"\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/cbook/__init__.py:1402: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" ndim = x[:, None].ndim\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:276: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" x = x[:, np.newaxis]\n",
"/home/sioni/miniconda3/envs/hls4ml-cnn/lib/python3.7/site-packages/matplotlib/axes/_base.py:278: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version. Convert to a numpy array before indexing instead.\n",
" y = y[:, np.newaxis]\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 648x648 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"do_a_plot_2('latency_us', 'Latency (μs)')"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>im_height</th>\n",
" <th>im_width</th>\n",
" <th>n_filt</th>\n",
" <th>filt_height</th>\n",
" <th>filt_width</th>\n",
" <th>dsp</th>\n",
" <th>lut</th>\n",
" <th>ff</th>\n",
" <th>bram</th>\n",
" <th>latency_clks</th>\n",
" <th>latency_us</th>\n",
" <th>error</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>10</td>\n",
" <td>10</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>27</td>\n",
" <td>1254</td>\n",
" <td>410</td>\n",
" <td>0.0</td>\n",
" <td>502</td>\n",
" <td>2.51</td>\n",
" <td>False</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>10</td>\n",
" <td>10</td>\n",
" <td>1</td>\n",
" <td>5</td>\n",
" <td>5</td>\n",
" <td>75</td>\n",
" <td>3174</td>\n",
" <td>838</td>\n",
" <td>0.5</td>\n",
" <td>602</td>\n",
" <td>3.01</td>\n",
" <td>False</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>10</td>\n",
" <td>10</td>\n",
" <td>1</td>\n",
" <td>7</td>\n",
" <td>7</td>\n",
" <td>148</td>\n",
" <td>6051</td>\n",
" <td>1425</td>\n",
" <td>1.0</td>\n",
" <td>702</td>\n",
" <td>3.51</td>\n",
" <td>False</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>60</td>\n",
" <td>60</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>27</td>\n",
" <td>1533</td>\n",
" <td>1400</td>\n",
" <td>13.5</td>\n",
" <td>18002</td>\n",
" <td>90.01</td>\n",
" <td>False</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>60</td>\n",
" <td>60</td>\n",
" <td>1</td>\n",
" <td>5</td>\n",
" <td>5</td>\n",
" <td>75</td>\n",
" <td>4286</td>\n",
" <td>3748</td>\n",
" <td>38.0</td>\n",
" <td>21602</td>\n",
" <td>108.00</td>\n",
" <td>False</td>\n",
" </tr>\n",
" <tr>\n",
" <th>...</th>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>655</th>\n",
" <td>249</td>\n",
" <td>249</td>\n",
" <td>64</td>\n",
" <td>5</td>\n",
" <td>5</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.00</td>\n",
" <td>True</td>\n",
" </tr>\n",
" <tr>\n",
" <th>656</th>\n",
" <td>249</td>\n",
" <td>249</td>\n",
" <td>64</td>\n",
" <td>7</td>\n",
" <td>7</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.00</td>\n",
" <td>True</td>\n",
" </tr>\n",
" <tr>\n",
" <th>657</th>\n",
" <td>256</td>\n",
" <td>256</td>\n",
" <td>64</td>\n",
" <td>3</td>\n",
" <td>3</td>\n",
" <td>1716</td>\n",
" <td>17440</td>\n",
" <td>4197</td>\n",
" <td>13.5</td>\n",
" <td>327682</td>\n",
" <td>1638.00</td>\n",
" <td>False</td>\n",
" </tr>\n",
" <tr>\n",
" <th>658</th>\n",
" <td>256</td>\n",
" <td>256</td>\n",
" <td>64</td>\n",
" <td>5</td>\n",
" <td>5</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.00</td>\n",
" <td>True</td>\n",
" </tr>\n",
" <tr>\n",
" <th>659</th>\n",
" <td>256</td>\n",
" <td>256</td>\n",
" <td>64</td>\n",
" <td>7</td>\n",
" <td>7</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.00</td>\n",
" <td>True</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>660 rows × 12 columns</p>\n",
"</div>"
],
"text/plain": [
" im_height im_width n_filt filt_height filt_width dsp lut ff \\\n",
"0 10 10 1 3 3 27 1254 410 \n",
"1 10 10 1 5 5 75 3174 838 \n",
"2 10 10 1 7 7 148 6051 1425 \n",
"3 60 60 1 3 3 27 1533 1400 \n",
"4 60 60 1 5 5 75 4286 3748 \n",
".. ... ... ... ... ... ... ... ... \n",
"655 249 249 64 5 5 0 0 0 \n",
"656 249 249 64 7 7 0 0 0 \n",
"657 256 256 64 3 3 1716 17440 4197 \n",
"658 256 256 64 5 5 0 0 0 \n",
"659 256 256 64 7 7 0 0 0 \n",
"\n",
" bram latency_clks latency_us error \n",
"0 0.0 502 2.51 False \n",
"1 0.5 602 3.01 False \n",
"2 1.0 702 3.51 False \n",
"3 13.5 18002 90.01 False \n",
"4 38.0 21602 108.00 False \n",
".. ... ... ... ... \n",
"655 0.0 0 0.00 True \n",
"656 0.0 0 0.00 True \n",
"657 13.5 327682 1638.00 False \n",
"658 0.0 0 0.00 True \n",
"659 0.0 0 0.00 True \n",
"\n",
"[660 rows x 12 columns]"
]
},
"execution_count": 24,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.10"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment