benman1

import tensorflow as tf
config=tf.ConfigProto(
    log_device_placement=True,
)
sess = tf.Session(config=config)
assert tf.test.is_gpu_available()
assert tf.test.is_built_with_cuda()
op = tf.shape(
    tf.nn.conv2d(tf.random_normal([1,10,10,10]),
                 tf.random_normal([2,10,10,10]),

benman1

import tensorflow as tf
from tensorflow.python.saved_model import tag_constants
from tensorflow.python.saved_model.signature_def_utils_impl import predict_signature_def

KERAS_HDF5 = 'weights.hdf5'
EXPORT_PATH = '/tmp/tf_model/'  # where to save the tensorflow model


model = create_model(params)  # build and compile the architecture
model.load_weights(KERAS_HDF5)

benman1

[alias]
    unadd = reset HEAD
    st = status
    co = checkout
    uncommit = reset --soft HEAD^
    changed = diff-tree -no-commit-id --name-only -r
    updatesubmodules = pull --recurse-submodules && git submodule update
    hist = log --pretty=format:\"%h %ad | %s%d [%an]\" --graph --date=short    
[pull]
    rebase = true

benman1

# USAGE:
# ./parallel.sh <command> <input directory> [<threads> [<log file>]]
#
# EXAMPLE:
# ./parallel.sh longrunningcommand.java /myfiles/input/ 10 process.log
shell_command=$1
input_dir=$2
threads=$3
threads=${threads:=10}
output_log=$4

benman1

# display the biggest files within a directory
find mypath -type f -printf '%s %p\n'| sort -nr | head -10

benman1

#!/bin/sh

find . -name '*.py' -exec dos2unix {} \;
find . -name '*.py' -exec 2to3 -w {} \;
find . -name '*.py' -exec autopep8 --in-place --aggressive --aggressive {} \;
find . -name '*.py' -exec docformatter --in-place {} \;
isort -rc --atomic .
find . -name '*.py' -exec vim -c '%s/except \(.*\), e:/except \1 as e:/' {} -c 'wq' \;
find . -name '*.py' -exec autoflake --in-place --remove-unused-variables {} \;
find . -name '*.py' -exec vim -c '%s/xrange/range/' {} -c 'wq' \;

benman1

import sys
import pandas as pd
from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
from sklearn.datasets import load_iris


def extract_bins(tree):
    '''Extract bins from tree
    '''
    edges = {}

benman1

adapted ftom from https://github.com/jonathansick/dotfiles/wiki/Notes-for-Mac-OS-X

ATLAS & LAPACK

ATLAS can be downloaded from SourceForge; these instructions are for atlas 3.10.3. Also grab a copy of LAPACK with

wget http://www.netlib.org/lapack/lapack-3.8.0.tar.gz

Setup Move ATLAS to ~/build. Unpack and setup via:

benman1

# found at https://stackoverflow.com/questions/6620471/fitting-empirical-distribution-to-theoretical-ones-with-scipy-python

%matplotlib inline

import warnings
import numpy as np
import pandas as pd
import scipy.stats as st
import statsmodels as sm
import matplotlib

benman1

import streams, nre

#[ This script reads a template file and substitutes according to
key-value pairs. All variables in the template start with $. For 
example $elementid.
]#


type
  Substitutions = seq[tuple[name: string, value: string]]