def predictOutput(image="DATA/mask/15_h.jpg",
weights='model_w_new_512.h5',
output_name='512result',
channel=1,
grey=False,
dim=(512, 512, 1),
f_ratio=[0.5]):
IMAGE_PATH = image
WEIGHTS_PATH = weights
OUTPUT_PATH = output_name
CHANNEL = channel
GREY = grey
model = net.NeuralNetwork()
extractor = de.DataExtractor()
model.load(dim, WEIGHTS_PATH)
data = extractor.extractData(IMAGE_PATH,
None,
channel=CHANNEL,
shape=dim,
grey=GREY)
#data=reframe(data,dim)
shape = extractor.shape2(IMAGE_PATH, shape=dim)
result = model.predictSet(data, shape, dim)
io.imsave(OUTPUT_PATH + '.jpg', result)
for ratio in f_ratio:
im = scipy.misc.toimage(extractor.cutOff(result, ratio))
im.save(OUTPUT_PATH + '_' + str(ratio) + '_filtered.png')
gc.collect()
def learnNetwork(image="DATA/mask/05_h.jpg",
mask="DATA/mask/05_h.tif",
mode='create',
old_weights='model_w_new_128.h5',
new_weights='model_w_new_128.h5',
epochs=1,
channel=1,
gray=False,
dim=(128, 128, 1)):
IMAGE_PATH = image
MASK_PATH = mask
WEIGHTS_PATH = old_weights
NEW_WEIGHTS_PATH = new_weights
EPOCH_NUMBER = epochs
CHANNEL = channel
GREY = gray
MODE = mode
model = net.NeuralNetwork()
extractor = de.DataExtractor()
if MODE == 'create':
model.create(dim)
elif MODE == 'learn':
model.load(dim, WEIGHTS_PATH)
else:
print('Bad mode. End of script...')
sys.exit()
train = extractor.extractData(IMAGE_PATH,
MASK_PATH,
channel=CHANNEL,
shape=dim,
grey=GREY)
model.learn(train, epoch=EPOCH_NUMBER)
model.save_weights(NEW_WEIGHTS_PATH)
del train
del model
gc.collect()
import DataExtractor import ConnectionManager dataExtractor = DataExtractor.DataExtractor() ConnectionManager.connManager.registerReceiver(dataExtractor)
"""
import gc
import numpy as np
import network as net
import DataExtractor as de
from skimage import io
dim = (8, 8, 1)
if dim[2] == 1:
g = True
else:
g = False
model = net.NeuralNetwork()
extractor = de.DataExtractor()
trainTemp = extractor.extractData("DATA/mask/01_h.jpg",
"DATA/mask/01_h.tif",
shape=dim,
grey=g)
train = (np.reshape(trainTemp[0],
(trainTemp[0].shape[0], dim[2], dim[0], dim[1])),
trainTemp[1])
del trainTemp
model.create(size=dim)
model.learn(train, epoch=10)
trainTemp = extractor.extractData("DATA/mask/02_h.jpg",
import sys import pandas as pd sys.path.insert(0, "function/") from DataExtractor import * from RegClf import * #data extraction data = DataExtractor() X_train, X_test, Y_train, Y_test = data.extract() #fit and choose are regressions regression = RegressionClassifier(X_train, X_test, Y_train, Y_test) #choose are regression types regression.defineClassifierType([2, 3, 4, 5, 6]) regression.fit() regression.predict() regression.score() bestReg = regression.mostAccurateScore() regression.plotGraph(data.getRawExtract())
def __init__(self, filename, columns=None, background=False, outputDirectory=None, startColumn=None, saveImages=True):
self.dataExtractor = DataExtractor.DataExtractor(filename)
self.plotDataWriter = PlotDataWriter.PlotDataWriter()
self.saveImages = saveImages
self.dataPlot = DataPlot.DataPlot(self.done, self.skip, self.dump, self.exit, self.prev, background)
self.background = background
availablePlotColumns = self.dataExtractor.getAvailableColumns('background' if background else 'plot')
try:
if outputDirectory == None:
filename = os.path.basename(filename) + '_results_' + str(int(time.time()))
self.outputDirectory = os.path.join(os.getcwd(), filename)
else:
self.outputDirectory = outputDirectory
# don't allow overwriting of directories
if os.path.exists(self.outputDirectory):
print("ERROR: Output directory already exists. Choose another directory or remove the exisiting output directory before continuing.")
sys.exit()
# attempt to make the output directory if it doesn't exist
if not os.path.exists(self.outputDirectory):
os.makedirs(self.outputDirectory)
# attempt to add the image directory if it's required
if saveImages:
self.imageDirectory = os.path.join(self.outputDirectory, 'plots')
if not os.path.exists(self.imageDirectory):
os.makedirs(self.imageDirectory)
print("Files will be output to: " + self.outputDirectory)
except Exception as e:
print("There was a problem creating the output directories.")
print(e)
sys.exit()
# availableBackgroundcolumns = self.dataExtractor.getAvailableColumns('background')
#
# backgroundDataWriter = BackgroundDataWriter.BackgroundDataWriter()
# for column in availableBackgroundcolumns:
# _, yData = self.dataExtractor.extractData(column)
# backgroundDataWriter.writeLine(column, sum(yData) / len(yData), np.var(yData))
# backgroundDataWriter.writeToFile(self.outputDirectory)
# print("Background data processing completed.")
#Ensure that the column names are valid
if columns != None:
errorColumns = []
for column in columns:
if not column in availablePlotColumns:
errorColumns.append(column)
if len(errorColumns) != 0:
print("Fatal Error! these column(s) do not exist for this mode: " + ', '.join(x for x in errorColumns))
sys.exit()
self.columnsToPlot = columns
elif startColumn != None:
try:
idx = availablePlotColumns.index(startColumn)
self.columnsToPlot = availablePlotColumns[idx:]
except:
print("Start column: " + startColumn + " was not found.")
else:
self.columnsToPlot = availablePlotColumns
if len(self.columnsToPlot) == 0:
print("No columns to plot. Exiting.")
sys.exit()
self.currentPlotIndex = 0
self.printGuide()
self.plotNext()
import learning, util, sys, DataExtractor
from learning import *
from DataExtractor import DataExtractor
learner = StochasticGradientLearner(footballFeatureExtractor)
data = list()
for i in xrange(5, 13):
data.append(DataExtractor(i).featureDictionary)
train = dict()
test = dict()
for i in xrange(0, 4):
train.update(data[i])
for i in xrange(4, 8):
test.update(data[i])
from optparse import OptionParser
parser = OptionParser()
def default(str):
return str + ' [Default: %default]'
parser.add_option(
'-f',
'--featureExtractor',
from DataExtractor import * data = DataExtractor(11) data10 = DataExtractor(10) data9 = DataExtractor(9) print len(data.featureDictionary.items()) print len(data10.featureDictionary.items()) data.featureDictionary.update(data10.featureDictionary) print len(data.featureDictionary.items())
