1. packing type 2 (0452a446) · Commits · Elphel / python3-imagej-tiff

pack_tile.py

+134 −3

Original line number	Original line	Diff line number	Diff line
	@@ -6,7 +6,8 @@ __email__ = "oleg@elphel.com"

	import numpy as np		import numpy as np

	# pack from 9x9x4 to 25x1		# pack from 9x9x4 to 100x1
			# type 1, all the same
	def pack_tile_type1(tile):		def pack_tile_type1(tile):

	out = np.empty(100)		out = np.empty(100)
	@@ -125,13 +126,143 @@ def pack_tile_type1(tile):

	return out		return out

			# pack from 9x9x4 to 104x1
			# type 1, all the same
			def pack_tile_type2(tile):

			out = np.empty(104)

			# pack diagm-pair (not tested)
			l = np.ravel(tile[:,:,0])
			out[ 0] = 1.0l[ 0]+1.0l[ 1]+1.0l[ 2]+1.0l[ 9]+1.0l[10]+1.0l[18]
			out[ 1] = 1.0l[ 3]+1.0l[11]+1.0l[19]+1.0l[27]
			out[ 2] = 1.0l[ 4]+1.0l[12]+1.0l[20]+1.0l[28]+1.0*l[36]
			out[ 3] = 1.0l[ 5]+1.0l[ 6]+1.0*l[14]
			out[ 4] = 1.0l[ 7]+1.0l[15]+1.0*l[23]
			out[ 5] = 1.0l[ 8]+1.0l[16]+1.0*l[24]
			out[ 6] = 1.0l[13]+1.0l[21]+1.0l[29]+1.0l[37]
			out[ 7] = 1.0l[17]+1.0l[25]+1.0*l[33]
			out[ 8] = 1.0l[22]+1.0l[30]+1.0*l[38]
			out[ 9] = 1.0l[26]+1.0l[34]+1.0*l[35]
			out[10] = 1.0*l[31]
			out[11] = 1.0*l[32]
			out[12] = 1.0*l[39]
			out[13] = 1.0*l[40]
			out[14] = 1.0*l[41]
			out[15] = 1.0l[42]+1.0l[50]+1.0*l[58]
			out[16] = 1.0l[43]+1.0l[51]+1.0l[59]+1.0l[67]
			out[17] = 1.0l[44]+1.0l[52]+1.0l[60]+1.0l[68]+1.0*l[76]
			out[18] = 1.0l[45]+1.0l[46]+1.0*l[54]
			out[19] = 1.0l[47]+1.0l[55]+1.0*l[63]
			out[20] = 1.0*l[48]
			out[21] = 1.0*l[49]
			out[22] = 1.0l[53]+1.0l[61]+1.0l[69]+1.0l[77]
			out[23] = 1.0l[56]+1.0l[64]+1.0*l[72]
			out[24] = 1.0l[57]+1.0l[65]+1.0*l[73]
			out[25] = 1.0l[62]+1.0l[70]+1.0l[71]+1.0l[78]+1.0l[79]+1.0l[80]
			out[26] = 1.0l[66]+1.0l[74]+1.0*l[75]

			# pack diago-pair (not tested)
			l = np.ravel(tile[:,:,1])
			out[27] = 1.0l[ 0]+1.0l[10]+1.0*l[20]
			out[28] = 1.0l[ 1]+1.0l[11]+1.0*l[21]
			out[29] = 1.0l[ 2]+1.0l[ 3]+1.0*l[12]
			out[30] = 1.0l[ 4]+1.0l[14]+1.0l[24]+1.0l[34]+1.0*l[44]
			out[31] = 1.0l[ 5]+1.0l[15]+1.0l[25]+1.0l[35]
			out[32] = 1.0l[ 6]+1.0l[ 7]+1.0l[ 8]+1.0l[16]+1.0l[17]+1.0l[26]
			out[33] = 1.0l[ 9]+1.0l[19]+1.0*l[29]
			out[34] = 1.0l[13]+1.0l[23]+1.0*l[43]
			out[35] = 1.0l[18]+1.0l[27]+1.0*l[28]
			out[36] = 1.0l[22]+1.0l[32]+1.0*l[42]
			out[37] = 1.0*l[30]
			out[38] = 1.0*l[31]
			out[39] = 1.0l[36]+1.0l[46]+1.0l[56]+1.0l[66]+1.0*l[76]
			out[40] = 1.0l[37]+1.0l[47]+1.0l[57]+1.0l[67]
			out[41] = 1.0l[38]+1.0l[48]+1.0*l[58]
			out[42] = 1.0*l[39]
			out[43] = 1.0*l[40]
			out[44] = 1.0*l[41]
			out[45] = 1.0l[45]+1.0l[55]+1.0l[65]+1.0l[75]
			out[46] = 1.0*l[49]
			out[47] = 1.0*l[50]
			out[48] = 1.0l[51]+1.0l[61]+1.0*l[71]
			out[49] = 1.0l[52]+1.0l[53]+1.0*l[62]
			out[50] = 1.0l[54]+1.0l[63]+1.0l[64]+1.0l[72]+1.0l[73]+1.0l[74]
			out[51] = 1.0l[59]+1.0l[69]+1.0*l[79]
			out[52] = 1.0l[60]+1.0l[70]+1.0*l[80]
			out[53] = 1.0l[68]+1.0l[77]+1.0*l[78]

			# pack hor-pairs
			l = np.ravel(tile[:,:,2])
			out[54] = 1.0l[ 0]+1.0l[ 1]+1.0l[ 9]+1.0l[10]+1.0l[18]+1.0l[27]+1.0l[36]+1.0l[45]+1.0l[54]+1.0l[63]+1.0l[64]+1.0l[72]+1.0*l[73]
			out[55] = 1.0l[ 2]+1.0l[11]+1.0*l[20]
			out[56] = 1.0l[ 3]+1.0l[12]+1.0*l[21]
			out[57] = 1.0l[ 4]+1.0l[13]+1.0*l[22]
			out[58] = 1.0l[ 5]+1.0l[14]+1.0*l[23]
			out[59] = 1.0l[ 6]+1.0l[15]+1.0*l[24]
			out[60] = 1.0l[ 7]+1.0l[ 8]+1.0l[16]+1.0l[17]+1.0l[26]+1.0l[35]+1.0l[44]+1.0l[53]+1.0l[62]+1.0l[70]+1.0l[71]+1.0l[79]+1.0*l[80]
			out[61] = 1.0l[19]+1.0l[28]+1.0l[37]+1.0l[46]+1.0*l[55]
			out[62] = 1.0l[25]+1.0l[34]+1.0l[43]+1.0l[52]+1.0*l[61]
			out[63] = 1.0l[29]+1.0l[38]+1.0*l[47]
			out[64] = 1.0*l[30]
			out[65] = 1.0*l[31]
			out[66] = 1.0*l[32]
			out[67] = 1.0l[33]+1.0l[42]+1.0*l[51]
			out[68] = 1.0*l[39]
			out[69] = 1.0*l[40]
			out[70] = 1.0*l[41]
			out[71] = 1.0*l[48]
			out[72] = 1.0*l[49]
			out[73] = 1.0*l[50]
			out[74] = 1.0l[56]+1.0l[65]+1.0*l[74]
			out[75] = 1.0l[57]+1.0l[66]+1.0*l[75]
			out[76] = 1.0l[58]+1.0l[67]+1.0*l[76]
			out[77] = 1.0l[59]+1.0l[68]+1.0*l[77]
			out[78] = 1.0l[60]+1.0l[69]+1.0*l[78]

			# pack vert-pairs
			l = np.ravel(tile[:,:,3])
			out[79] = 1.0l[ 0]+1.0l[ 1]+1.0l[ 2]+1.0l[ 3]+1.0l[ 4]+1.0l[ 5]+1.0l[ 6]+1.0l[ 7]+1.0l[ 8]+1.0l[ 9]+1.0l[10]+1.0l[11]+1.0l[16]+1.0l[17]
			out[80] = 1.0l[11]+1.0l[12]+1.0l[13]+1.0l[14]+1.0*l[15]
			out[81] = 1.0l[18]+1.0l[19]+1.0*l[20]
			out[82] = 1.0l[21]+1.0l[22]+1.0*l[23]
			out[83] = 1.0l[24]+1.0l[25]+1.0*l[26]
			out[84] = 1.0l[27]+1.0l[28]+1.0*l[29]
			out[85] = 1.0*l[30]
			out[86] = 1.0*l[31]
			out[87] = 1.0*l[32]
			out[88] = 1.0l[33]+1.0l[34]+1.0*l[35]
			out[89] = 1.0l[36]+1.0l[37]+1.0*l[38]
			out[90] = 1.0*l[39]
			out[91] = 1.0*l[40]
			out[92] = 1.0*l[41]
			out[93] = 1.0l[42]+1.0l[43]+1.0*l[44]
			out[94] = 1.0l[45]+1.0l[46]+1.0*l[47]
			out[95] = 1.0*l[48]
			out[96] = 1.0*l[49]
			out[97] = 1.0*l[50]
			out[98] = 1.0l[51]+1.0l[52]+1.0*l[53]
			out[99] = 1.0l[54]+1.0l[55]+1.0*l[56]
			out[100] = 1.0l[57]+1.0l[58]+1.0*l[59]
			out[101] = 1.0l[60]+1.0l[61]+1.0*l[62]
			out[102] = 1.0l[63]+1.0l[64]+1.0l[70]+1.0l[71]+1.0l[72]+1.0l[73]+1.0l[74]+1.0l[75]+1.0l[76]+1.0l[77]+1.0l[78]+1.0l[79]+1.0*l[80]
			out[103] = 1.0l[65]+1.0l[66]+1.0l[67]+1.0l[68]+1.0*l[69]

			return out

	# pack single		# pack single
	def pack_tile(tile):		def pack_tile(tile):
	return pack_tile_type1(tile)		return pack_tile_type1(tile)

	# pack all tiles		# pack all tiles
	def pack(tiles):		def pack(tiles,ptype=1):
	output = np.array([[pack_tile(tiles[i,j]) for j in range(tiles.shape[1])] for i in range(tiles.shape[0])])
			if ptype==1:
			pack_func = pack_tile_type1
			elif ptype==2:
			pack_func = pack_tile_type2

			output = np.array([[pack_func(tiles[i,j]) for j in range(tiles.shape[1])] for i in range(tiles.shape[0])])
	return output		return output

test_nn_feed.py

+120 −104

Original line number	Original line	Diff line number	Diff line
	@@ -46,6 +46,7 @@ def print_time():
	VALUES_LAYER_NAME = 'other'		VALUES_LAYER_NAME = 'other'
	LAYERS_OF_INTEREST = ['diagm-pair', 'diago-pair', 'hor-pairs', 'vert-pairs']		LAYERS_OF_INTEREST = ['diagm-pair', 'diago-pair', 'hor-pairs', 'vert-pairs']
	RADIUS = 1		RADIUS = 1
			TILE_PACKING_TYPE = 1

	DEBUG_PLT_LOSS = True		DEBUG_PLT_LOSS = True
	# If false - will not pack or rescal		# If false - will not pack or rescal
	@@ -91,22 +92,26 @@ def lrelu(x):

	def network(input):		def network(input):

	fc1 = slim.fully_connected(input,512,activation_fn=lrelu,scope='g_fc1')		fc1 = slim.fully_connected(input,1024,activation_fn=lrelu,scope='g_fc1')
	fc2 = slim.fully_connected(fc1, 2,activation_fn=lrelu,scope='g_fc2')		#fc2 = slim.fully_connected(fc1, 2,activation_fn=lrelu,scope='g_fc2')
	return fc2		#return fc2

	#fc2 = slim.fully_connected(fc1, 1024,activation_fn=lrelu,scope='g_fc2')		fc2 = slim.fully_connected(fc1, 1024,activation_fn=lrelu,scope='g_fc2')
	#fc3 = slim.fully_connected(fc2, 512,activation_fn=lrelu,scope='g_fc3')		fc3 = slim.fully_connected(fc2, 512,activation_fn=lrelu,scope='g_fc3')
	#fc4 = slim.fully_connected(fc3, 8,activation_fn=lrelu,scope='g_fc4')		fc4 = slim.fully_connected(fc3, 8,activation_fn=lrelu,scope='g_fc4')
	#fc5 = slim.fully_connected(fc4, 4,activation_fn=lrelu,scope='g_fc5')		fc5 = slim.fully_connected(fc4, 4,activation_fn=lrelu,scope='g_fc5')
	#fc6 = slim.fully_connected(fc5, 2,activation_fn=lrelu,scope='g_fc6')		fc6 = slim.fully_connected(fc5, 2,activation_fn=lrelu,scope='g_fc6')

	#return fc6		return fc6


	sess = tf.Session()		sess = tf.Session()

			if TILE_PACKING_TYPE==1:
	in_tile = tf.placeholder(tf.float32,[None,101])		in_tile = tf.placeholder(tf.float32,[None,101])
			elif TILE_PACKING_TYPE==2:
			in_tile = tf.placeholder(tf.float32,[None,105])

	gt = tf.placeholder(tf.float32,[None,2])		gt = tf.placeholder(tf.float32,[None,2])


	@@ -135,23 +140,23 @@ cf_w_norm = tf.nn.softmax(cf_w)

	#G_loss = tf.reduce_mean(tf.abs(tf.nn.softmax(out[:,1])out[:,0]-cf_w_normgt[:,0]))		#G_loss = tf.reduce_mean(tf.abs(tf.nn.softmax(out[:,1])out[:,0]-cf_w_normgt[:,0]))
	#G_loss = tf.reduce_mean(tf.squared_difference(out[:,0], gt[:,0]))		#G_loss = tf.reduce_mean(tf.squared_difference(out[:,0], gt[:,0]))
	#G_loss = tf.reduce_mean(tf.abs(out[:,0]-gt[:,0]))		G_loss = tf.reduce_mean(tf.abs(out[:,0]-gt[:,0]))
	G_loss = tf.losses.mean_squared_error(gt[:,0],out[:,0],cf_w)		#G_loss = tf.losses.mean_squared_error(gt[:,0],out[:,0],cf_w)

	tf.summary.scalar('loss', G_loss)		#tf.summary.scalar('loss', G_loss)
	tf.summary.scalar('prediction', out[0,0])		#tf.summary.scalar('prediction', out[0,0])
	tf.summary.scalar('ground truth', gt[0,0])		#tf.summary.scalar('ground truth', gt[0,0])

	t_vars=tf.trainable_variables()		t_vars=tf.trainable_variables()
	lr=tf.placeholder(tf.float32)		lr=tf.placeholder(tf.float32)
	G_opt=tf.train.AdamOptimizer(learning_rate=lr).minimize(G_loss,var_list=[var for var in t_vars if var.name.startswith('g_')])		G_opt=tf.train.AdamOptimizer(learning_rate=lr).minimize(G_loss)

	saver=tf.train.Saver()		saver=tf.train.Saver()

	# ?!!!!!		# ?!!!!!
	merged = tf.summary.merge_all()		#merged = tf.summary.merge_all()
	train_writer = tf.summary.FileWriter(result_dir + '/train', sess.graph)		#train_writer = tf.summary.FileWriter(result_dir + '/train', sess.graph)
	test_writer = tf.summary.FileWriter(result_dir + '/test')		#test_writer = tf.summary.FileWriter(result_dir + '/test')

	sess.run(tf.global_variables_initializer())		sess.run(tf.global_variables_initializer())
	ckpt=tf.train.get_checkpoint_state(checkpoint_dir)		ckpt=tf.train.get_checkpoint_state(checkpoint_dir)
	@@ -185,75 +190,90 @@ if DEBUG_PLT_LOSS:
	pass		pass


			training_tiles = np.array([])
			training_values = np.array([])

			# get epoch train data
			for i in range(len(tlist)):

	# RUN		print(bcolors.OKGREEN+"Opening "+tlist[i]+bcolors.ENDC)
	# epoch is one image


			tmp_tiff = ijt.imagej_tiff(tlist[i])
			tmp_tiles = tmp_tiff.getstack(labels,shape_as_tiles=True)
			tmp_vals = tmp_tiff.getvalues(label=VALUES_LAYER_NAME)
			# might not need it because going to loop through anyway
			if TILE_PACKING_TYPE==1:
			packed_tiles = pile.pack(tmp_tiles)
			elif TILE_PACKING_TYPE==2:
			packed_tiles = pile.pack(tmp_tiles,TILE_PACKING_TYPE)

	for epoch in range(lastepoch,lastepoch+len(tlist)):		packed_tiles = np.dstack((packed_tiles,tmp_vals[:,:,0]))

	print(bcolors.HEADER+"Epoch #"+str(epoch)+bcolors.ENDC)

	if os.path.isdir("result/%04d"%epoch):		packed_tiles = np.reshape(packed_tiles,(-1,packed_tiles.shape[-1]))
	continue		values = np.reshape(tmp_vals[:,:,1:3],(-1,2))

	tlist_index = epoch - lastepoch		packed_tiles_filtered = np.array([])

	print(bcolors.OKGREEN+"Processing "+tlist[tlist_index]+bcolors.ENDC)		print("Unfiltered: "+str(packed_tiles.shape))

	tmp_tiff = ijt.imagej_tiff(tlist[tlist_index])		for j in range(packed_tiles.shape[0]):
	tmp_tiles = tmp_tiff.getstack(labels,shape_as_tiles=True)
	tmp_vals = tmp_tiff.getvalues(label=VALUES_LAYER_NAME)

	# might not need it because going to loop through anyway		skip_tile = False
	packed_tiles = pile.pack(tmp_tiles)		if np.isnan(np.sum(packed_tiles[j])):
	packed_tiles = np.dstack((packed_tiles,tmp_vals[:,:,0]))		skip_tile = True
			if np.isnan(np.sum(values[j])):
			skip_tile = True

	#if epoch > 2000:		if not skip_tile:
	# LR = 1e-5		if len(packed_tiles_filtered)==0:
			packed_tiles_filtered = np.array([packed_tiles[j]])
			values_filtered = np.array([values[j]])
			else:
			packed_tiles_filtered = np.append(packed_tiles_filtered,[packed_tiles[j]],axis=0)
			values_filtered = np.append(values_filtered,[values[j]],axis=0)

	# so, here get the image, remove nans and run for 100x times		print("NaN-filtered: "+str(packed_tiles_filtered.shape))
	packed_tiles[np.isnan(packed_tiles)] = 0.0
	tmp_vals[np.isnan(tmp_vals)] = 0.0

	#packed_tiles = packed_tiles[::,::]		if i==0:
	values = tmp_vals		training_tiles = packed_tiles_filtered
			training_values = values_filtered
			else:
			training_tiles = np.concatenate((training_tiles,packed_tiles_filtered),axis=0)
			training_values = np.concatenate((training_values,values_filtered),axis=0)

	input_patch = np.reshape(packed_tiles,(-1,101))		print("Training set shape: "+str(training_tiles.shape))
	gt_patch = np.reshape(values[:,:,1:3],(-1,2))

	g_loss = np.zeros(input_patch.shape[0])		# RUN
			# epoch is all available images
			# batch is a number of non-zero tiles

			g_loss = np.zeros(training_tiles.shape[0])

	for i in range(100):		#for epoch in range(lastepoch,lastepoch+len(tlist)):
			for epoch in range(lastepoch,500):

	print(bcolors.OKBLUE+"Iteration "+str(i)+bcolors.ENDC)		print(bcolors.HEADER+"Epoch #"+str(epoch)+bcolors.ENDC)

	st=time.time()		if os.path.isdir("result/%04d"%epoch):
			continue

	skip_iteration = False		#if epoch > 2000:
			# LR = 1e-5

	# if nan skip run!		# so, here get the image, remove nans and run for 100x times
	if np.isnan(np.sum(gt_patch)):		#packed_tiles[np.isnan(packed_tiles)] = 0.0
	print("GT has NaNs")		#tmp_vals[np.isnan(tmp_vals)] = 0.0
	#skip_iteration = True

	if np.isnan(np.sum(input_patch)):		input_patch = training_tiles
	print("Patch has NaNs")		gt_patch = training_values
	#skip_iteration = True

	if skip_iteration:		st=time.time()
	#print(bcolors.WARNING+"Found NaN, skipping iteration for tile "+str(i)+","+str(j)+bcolors.ENDC)
	pass
	else:

	run_options = tf.RunOptions(trace_level=tf.RunOptions.FULL_TRACE)		#run_options = tf.RunOptions(trace_level=tf.RunOptions.FULL_TRACE)
	run_metadata = tf.RunMetadata()		#run_metadata = tf.RunMetadata()
			#_,G_current,output = sess.run([G_opt,G_loss,out],feed_dict={in_tile:input_patch,gt:gt_patch,lr:LR},options=run_options,run_metadata=run_metadata)

	_,G_current,output,summary = sess.run([G_opt,G_loss,out,merged],feed_dict={in_tile:input_patch,gt:gt_patch,lr:LR},options=run_options,run_metadata=run_metadata)		_,G_current,output = sess.run([G_opt,G_loss,out],feed_dict={in_tile:input_patch,gt:gt_patch,lr:LR})
	#_,G_current,output = sess.run([G_opt,G_loss,out],feed_dict={in_tile:input_patch,gt:gt_patch,lr:LR})

	g_loss[i]=G_current		g_loss[i]=G_current
	mean_loss = np.mean(g_loss[np.where(g_loss)])		mean_loss = np.mean(g_loss[np.where(g_loss)])
	@@ -300,17 +320,13 @@ for epoch in range(lastepoch,lastepoch+len(tlist)):

	else:		else:
	print("%d %d Loss=%.3f CurrentLoss=%.3f Time=%.3f"%(epoch,i,mean_loss,G_current,time.time()-st))		print("%d %d Loss=%.3f CurrentLoss=%.3f Time=%.3f"%(epoch,i,mean_loss,G_current,time.time()-st))
	#train_writer.add_run_metadata(run_metadata, 'step%d' % cnt)
	#test_writer.add_summary(summary,cnt)
	#train_writer.add_summary(summary, cnt)

	if epoch%save_freq==0:		if epoch%save_freq==0:
	if not os.path.isdir(result_dir + '%04d'%epoch):		if not os.path.isdir(result_dir + '%04d'%epoch):
	os.makedirs(result_dir + '%04d'%epoch)		os.makedirs(result_dir + '%04d'%epoch)

	saver.save(sess, checkpoint_dir + 'model.ckpt')		saver.save(sess, checkpoint_dir + 'model.ckpt')
	train_writer.close()
	test_writer.close()

	print_time()		print_time()
	print(bcolors.OKGREEN+"time: "+str(time.time())+bcolors.ENDC)		print(bcolors.OKGREEN+"time: "+str(time.time())+bcolors.ENDC)

test_nn_infer.py

+22 −3

Original line number	Original line	Diff line number	Diff line
	@@ -44,6 +44,7 @@ def print_time():
	VALUES_LAYER_NAME = 'other'		VALUES_LAYER_NAME = 'other'
	LAYERS_OF_INTEREST = ['diagm-pair', 'diago-pair', 'hor-pairs', 'vert-pairs']		LAYERS_OF_INTEREST = ['diagm-pair', 'diago-pair', 'hor-pairs', 'vert-pairs']
	RADIUS = 1		RADIUS = 1
			TILE_PACKING_TYPE = 1

	try:		try:
	src = sys.argv[1]		src = sys.argv[1]
	@@ -83,7 +84,11 @@ def network(input):

	sess = tf.Session()		sess = tf.Session()

			if TILE_PACKING_TYPE==1:
	in_tile = tf.placeholder(tf.float32,[None,101])		in_tile = tf.placeholder(tf.float32,[None,101])
			elif TILE_PACKING_TYPE==2:
			in_tile = tf.placeholder(tf.float32,[None,105])

	gt = tf.placeholder(tf.float32,[None,2])		gt = tf.placeholder(tf.float32,[None,2])
	out = network(in_tile)		out = network(in_tile)

	@@ -143,7 +148,11 @@ for item in tlist:
	# tiles and values		# tiles and values

	# might not need it because going to loop through anyway		# might not need it because going to loop through anyway
			if TILE_PACKING_TYPE==1:
	packed_tiles = pile.pack(tiles)		packed_tiles = pile.pack(tiles)
			elif TILE_PACKING_TYPE==2:
			packed_tiles = pile.pack(tiles,TILE_PACKING_TYPE)

	packed_tiles = np.dstack((packed_tiles,values[:,:,0]))		packed_tiles = np.dstack((packed_tiles,values[:,:,0]))

	print(packed_tiles.shape)		print(packed_tiles.shape)
	@@ -168,11 +177,12 @@ for item in tlist:
	packed_tiles_flat = packed_tiles[i]		packed_tiles_flat = packed_tiles[i]
	values_flat = values[i]		values_flat = values[i]

			# whole row at once
	output = sess.run(out,feed_dict={in_tile:packed_tiles_flat})		output = sess.run(out,feed_dict={in_tile:packed_tiles_flat})
	output_image[i] = output		output_image[i] = output

	# so, let's print		# so, let's print
	for j in range(output.shape[0]):		for j in range(packed_tiles.shape[0]):
	p = output[j,0]		p = output[j,0]
	pc = output[j,1]		pc = output[j,1]
	fv = values_flat[j,0]		fv = values_flat[j,0]
	@@ -204,6 +214,15 @@ for item in tlist:

	tif = np.dstack((im1,im2,im3))		tif = np.dstack((im1,im2,im3))

			im3 = np.ravel(im3)

			print(im3.shape)

			im4 = im3[~np.isnan(im3)]

			rms = np.sqrt(np.mean(np.square(im4)))
			print("RMS = "+str(rms))

	imagej_tiffwriter.save('prediction_results.tiff',tif)		imagej_tiffwriter.save('prediction_results.tiff',tif)

	#sys.exit(0)		#sys.exit(0)

Admin message