util.py

import time, sys, os, random
import numpy as np
import tensorflow as tf

def timestamp():
	return time.strftime( "%Y%m%d%H%M%S", time.gmtime() )
#end timestamp

def memory_usage():
	pid=os.getpid()
	s = next( line for line in open( '/proc/{}/status'.format( pid ) ).read().splitlines() if line.startswith( 'VmSize' ) ).split()
	return "{} {}".format( s[-2], s[-1] )
#end memory_usage

def sparse_to_dense( M_sparse, default = 0.0 ):
	M_i, M_v, M_shape = M_sparse
	n, m = M_shape
	M = np.ones( (n, m), dtype = np.float32 ) * default
	for indexes, value in zip( M_i, M_v ):
		i,j = indexes
		M[i,j] = value
	#end for
	return M
#end sparse_to_dense

def dense_to_sparse( M, check = lambda x: x != 0, val = lambda x: x ):
	n, m = M.shape
	M_i = []
	M_v = []
	M_shape = (n,m)
	for i in range( n ):
		for j in range( m ):
			if check( M[i,j] ):
				M_i.append( (i,j ) )
				M_v.append( val( M[i,j] ) )
			#end if
		#end for
	#end for
	return (M_i,M_v,M_shape)
#end dense_to_sparse

def reindex_matrix( n, m, M ):
	new_index = []
	new_value = []
	for i, v in zip( M[0], M[1] ):
		s, t = i
		new_index.append( (n + s, m + t) )
		new_value.append( v )
	#end for
	return zip( new_index, new_value )
#end reindex_matrix

def load_weights(sess,path,scope=None):
	if os.path.exists(path):
		# Restore saved weights
		print( "{timestamp}\t{memory}\tRestoring saved model ... ".format( timestamp = timestamp(), memory = memory_usage() ) )
		# Create model saver
		if scope is None:
			saver = tf.train.Saver()
		else:
			saver = tf.train.Saver(var_list=tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope=scope))
		#end
		saver.restore(sess, "%s/model.ckpt" % path)
	#end if
#end

def save_weights(sess,path,scope=None):
	# Create /tmp/ directory to save weights
	if not os.path.exists(path):
		os.makedirs(path)
	#end if
	# Create model saver
	if scope is None:
		saver = tf.train.Saver()
	else:
		saver = tf.train.Saver(var_list=tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES, scope=scope))
	#end
	saver.save(sess, "%s/model.ckpt" % path)
	print( "{timestamp}\t{memory}\tMODEL SAVED IN PATH: {path}".format( timestamp = timestamp(), memory = memory_usage(), path=path ) )
#end