sbalnojan

import s3fs
import fire
import os


class S3CopyMachine(object):
  """Copy to S3 via s3fs."""

  def to_s3(self, local_bucket, s3_bucket):

sbalnojan

import s3fs
import pickle
import json
import numpy as np

BUCKET_NAME = "my-bucket"

# definitions, keras/tf/... imports...

if __name__ == "__main__":

sbalnojan

# ...
# your data preprocessing...
if __name__ == "__main__":

  # ...
  # your data preprocessing...

  s3 = s3fs.S3FileSystem(anon=False)
  df = pd.DataFrame(data={"foo":[0]})
  print(df)

sbalnojan

import os, sys
sys.path.append(os.path.join(os.getcwd(), "keras-deep-graph-learning")) # Adding the submodule to the module search path
sys.path.append(os.path.join(os.getcwd(), "keras-deep-graph-learning/examples")) # Adding the submodule to the module search path
import numpy as np
from keras.layers import Dense, Activation, Dropout
from keras.models import Model, Sequential
from keras.regularizers import l2
from keras.optimizers import Adam
from keras_dgl.layers import GraphCNN
import keras.backend as K

sbalnojan

train_on_weight= np.array([1,1,0])
print("Now we won't do any fancy preprocessing, just basic training.")

NUM_FILTERS = 1
graph_conv_filters =  A # you may try np.eye(3)
graph_conv_filters = K.constant(graph_conv_filters)

model = Sequential()
model.add(GraphCNN(Y.shape[1], NUM_FILTERS, graph_conv_filters, input_shape=(X.shape[1],), activation='elu', kernel_regularizer=l2(5e-4)))
model.add(Activation('softmax'))

sbalnojan

import os, sys
sys.path.append(os.path.join(os.getcwd(), "keras-deep-graph-learning")) # Adding the submodule to the module search path
sys.path.append(os.path.join(os.getcwd(), "keras-deep-graph-learning/examples")) # Adding the submodule to the module search path
import numpy as np
from examples import utils
from keras.layers import Dense, Activation, Dropout
from keras.models import Model, Sequential
from keras.regularizers import l2
from keras.optimizers import Adam
from keras_dgl.layers import GraphCNN

sbalnojan

 print("this simply ignores the connected edges, thus receives a pretty bad test_acc:")
graph_conv_filters = np.eye(A_norm.shape[0])
graph_conv_filters = K.constant(graph_conv_filters)
num_filters = 1

model = Sequential()
model.add(GraphCNN(Y.shape[1], num_filters, graph_conv_filters, input_shape=(X.shape[1],), activation='elu',
                       kernel_regularizer=l2(5e-4)))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy', optimizer=Adam(lr=0.01), metrics=['acc'])

sbalnojan

print("this filter includes the edges, so it should perform considerably better than before.:")
graph_conv_filters = A_norm
graph_conv_filters = K.constant(graph_conv_filters)
num_filters = 1

model = Sequential()
model.add(GraphCNN(Y.shape[1], num_filters, graph_conv_filters, input_shape=(X.shape[1],), activation='elu',
                       kernel_regularizer=l2(5e-4)))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy', optimizer=Adam(lr=0.01), metrics=['acc'])

sbalnojan

graph_conv_filters = np.concatenate([A_norm, np.matmul(A_norm, A_norm)], axis=0)
graph_conv_filters = K.constant(graph_conv_filters)
num_filters = 2

model = Sequential()
model.add(GraphCNN(Y.shape[1], num_filters, graph_conv_filters, input_shape=(X.shape[1],), activation='elu',
                       kernel_regularizer=l2(5e-4)))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy', optimizer=Adam(lr=0.01), metrics=['acc'])
model.summary()

sbalnojan

import os
import random

def convert_path(fname):
    basename, _ = os.path.splitext(fname)
    out_dir = basename + '_partitioned'
    return out_dir

def random_split_file(fpath):
    root = os.path.dirname(fpath)