def case_function(samples, obj):
sam = Samples() # object required to plot
num_bin_x, num_bin_y = (bin_x.get(), bin_y.get())
mi_text.set(mi_text.get() + '\nCalculating...')
not_selected = True
if plot_sample.get():
sam.plot_sample(samples, num_bin_x, num_bin_y)
not_selected = False
if plot_compare.get():
obj.plot_compare2(num_bin_x, num_bin_y, noise.get(), step.get(), significance.get())
not_selected = False
if plot_proposal.get():
sam.plot_propose(samples, num_bin_x, step.get())
not_selected = False
if plot_partition.get():
sam.plot_partition(samples, num_bin_x, num_bin_y)
not_selected = False
if plot_converge_in_samples.get():
converge_in_samples(num_bin_x, num_bin_y, obj)
not_selected = False
if samples_check.get():
samples_grid(num_bin_x, num_bin_y)
not_selected = False
if bin_selection.get():
bin_selection_validation(sigma=significance.get(), obj=obj)
not_selected = False
if cal_measures.get():
calculate_measures(samples, num_bin_x, num_bin_y)
not_selected = False
if not_selected:
mi_text.set('Select a plot or calculation')
def case_function(samples, obj):
sam = Samples()
print(samples)
numBinx, numBiny = Binx.get(), Biny.get()
#numBinx = 30
#numBiny = 30
print(numBinx, numBiny)
print(numBinx)
x, y = samples[:, 0], samples[:, 1]
MI = mutual_info(x, y, numBinx, numBiny)
ex = entropy(x, numBinx)
ey = entropy(y, numBiny)
R = norm_MI(x, y, numBinx, numBiny)
rxy = pearson_corr(x, y)
exy = entropyx_y(x, y, numBinx, numBiny)
eyx = entropyx_y(y, x, numBinx, numBiny)
MIE = MI_Entropy(x, y, numBinx, numBiny)
Ixy = propuesta_Ixy(x, y, numBinx, numBiny)
Iyx = propuesta_Iyx(x, y, numBinx, numBiny)
Ixy2 = propuesta2_Ixy(x, y, numBinx, numBiny)
Iyx2 = propuesta2_Iyx(x, y, numBinx, numBiny)
PMD = propuesta_mutual_dependency(x, y, numBinx, numBiny)
PMD2 = propuesta2_mutual_dependency(x, y, numBinx, numBiny)
# d_cor = d_corr(x,y)
# Mic = MIC(x,y)
mitexto.set(
'Entropy of x: ' + str(ex) + '\n' + 'Entropy of y: ' + str(ey) + '\n' +
'Mutual information: ' + str(MI) + '\n' +
'Mutual info with entropy: ' + str(MIE) + '\n' + '\n' +
'(Max=1)Normalized Mutual info: ' + str(R) + '\n' +
'(Max=1)Pearson Correlation: ' + str(rxy) + '\n' +
# '(Max=1)Distance Correlation: ' + str(d_cor) + '\n' +
# '(Max=1)MIC: ' + str(Mic) + '\n' +
'(Max=1)Mutual Dependency: ' + str(PMD) + '\n' +
'(Max=1)Mutual Dependency2: ' + str(PMD2) + '\n' + '\n' +
'Entropy of X|Y = ' + str(exy) + '\n' + 'Entropy of Y|X = ' +
str(eyx) + '\n' + '(Max=1)Information in Y of X: ' + str(Ixy) + '\n' +
'(Max=1)Information in X of Y: ' + str(Iyx) + '\n' +
'(Max=1)Information2 in Y of X: ' + str(Ixy2) + '\n' +
'(Max=1)Information2 in X of Y: ' + str(Iyx2))
if plot_sample.get():
sam.plot_sample(samples, numBinx, numBiny, step.get())
if plot_compare.get():
obj.plot_compare2(numBinx, numBiny, noise.get(), step.get())
if plot_propuse.get():
sam.plot_propose(samples, numBinx)
if plot_partition.get():
sam.plot_partition(samples, numBinx, numBiny)
if plot_converge_in_samples.get():
Converge_in_samples(numBinx, numBiny, obj)
if samples_check.get():
samples_grid(numBinx, numBiny)
def save_wav(self) -> None:
# Save waveform
duration, ok = QtGui.QInputDialog.getInt(self, "Seconds of Audio:",
"Seconds:", 1, 0, MAX_VAL,
STEP_VAL)
if ok:
dialog = QtGui.QFileDialog()
path = dialog.getSaveFileName(self, 'Save File', os.getenv('HOME'),
'WAV (*.wav)')
if path[0] != '':
samples = Samples()
expression = copy.deepcopy(self.expression)
samples.set_expression(expression)
samples.gen_write_16(path[0], duration)
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.setup_layout()
self.setup_waveform()
self.setup_instructions()
self.layout.nextRow()
self.setup_spectrograph()
self.layout.nextRow()
self.setup_editor()
self.setup_index()
self.samples = Samples(self.waveform.data_available,
self.spectrograph.data_available)
self.copy_func_to_samples()
self.copy_func_to_editor_and_display()
self.setup_pyaudio()
self.stream.start_stream()
def __init__(self, json_path, out_dir):
# Attributes #
self.out_dir = out_dir
self.json_path = json_path
# Parse #
with open(json_path) as handle: self.info = json.load(handle)
# Basic #
self.account = self.info['uppmax_id']
self.run_num = self.info['run_num']
self.run_label = self.info['run_id']
self.project_short_name = self.info['project']
self.project_long_name = self.info['project_name']
# Own attributes #
self.num = self.info['pool_num']
self.label = self.info['pool_id']
self.short_name = self.info['pool']
self.long_name = self.info['pool_name']
self.id_name = "run%03d-pool%02d" % (self.run_num, self.num)
# Special #
self.samples = Samples(self)
self.primers = TwoPrimers(self)
# Second init #
self.loaded = False
def command_parser():
'''
THIS PART CAN PARSER THE COMMAND PARAMETERS.
Value 0 is off, 1 is open.
'''
try:
source_path = ""
save_path = ""
# The default value of data_deal model.
iamges_deal = 0
size_x = 28
size_y = 28
channel = 3
# samples model parameters
samples = 0
no_replace = 0
replace = 0
number = 10000
# train and test the model
train_test = 0
dataset = 'mnist.pkl.gz'
config_path = "../config-example"
n_y = 10
m1 = 0
m2 = 0
opts, args = getopt.getopt(sys.argv[1:], "ho:", \
["help", \
"images_deal", \
"images_path=", \
"size_x=", \
"size_y=" , \
"channel=", \
"samples", \
"no_replace=", \
"replace=", \
"source_path=" , \
"number=", \
"save_path=", \
"train_test", \
"dataset=", \
"config_path=", \
"n_y=" , \
"m1=", \
"m2="
])
#print opts,args
for x, y in opts:
if x in ("-h", "--help"):
print "-----------------------------------------------------------"
print "ensemb: command line brew"
print "usage : python ensemb.py <command> <args>"
print "---------------------------------------------------------------------"
print "images_deal : deal the image to pickle file"
print "---------- ensemb.py --images_deal "
print "---------- --source_path=/home/gd/ensemble-nn/data/image"
print "---------- --size_x=28"
print "---------- --size_y=28 "
print "---------- --channel=3"
print "---------------------------------------------------------------------"
print "samples : sample the samples from the datasets"
print "---------- ensemb.py --samples "
print "---------- --no_replace=1"
print "---------- --replace=0 "
print "---------- --source_path=./data/mnist.pkl.gz"
print "---------- --number=30000"
print "---------- --save_path=./data/mnist-sample-replace.pkl"
print "---------------------------------------------------------------------"
print "train_test : train and test the model"
print "---------- ensemb.py --train_test "
print "---------- --dataset =/home/gd/ensemble-nn/data/mnist.pkl.gz"
print "---------- --config_path=./config-example "
print "---------- --n_y =10"
print "---------- --m1 = 1"
print "---------------------------------------------------------------------"
# MODEL IMAGES_DEAL
if x in ("--images_deal"):
print "Deal the iamge .."
iamges_deal = 1
if x in ("--source_path"):
images_path = y
if x in ("--size_x"):
size_x = int(y)
if x in ("--size_y"):
size_y = int(y)
if x in ("--channel"):
channel = int(y)
# MODEL SAMPLES
if x in ("--samples"):
print "sampling .."
samples = 1
if x in ("--no_replace"):
no_replace = int(y)
if x in ("--replace"):
replace = int(y)
if x in ("--source_path"):
source_path = y
if x in ("--number"):
number = int(y)
if x in ("--save_path"):
save_path = y
# TRAIN AND TEST A MODEL
if x in ("--train_test"):
print "train and test .."
train_test = 1
if x in ("--dataset"):
dataset = y
if x in ("--config_path"):
config_path = y
if x in ("--n_y"):
n_y = int(y)
if x in ("--m1"):
m1 = int(y)
if x in ("--m2"):
m2 = int(y)
if iamges_deal == 1:
source_path = "/home/gd/ensemble-nn/data/image"
DealDate_deal(image_path=source_path,
size_x=size_x,
size_y=size_y,
channel=channel)
if samples == 1:
source_path = "/home/gd/ensemble-nn/data/mnist.pkl.gz"
save_path = "/home/gd/ensemble-nn/data/mnist-sample-replace.pkl"
if no_replace == 0 and replace == 0:
print "no_replace or replace ?"
if no_replace == 1 and replace == 1:
print "no_replace or replace ?"
if no_replace == 1:
instance = Samples()
instance.no_replace_sample(source_path, number, save_path)
if replace == 1:
instance = Samples.Samples()
instance.replace_sample(source_path, number, save_path)
if train_test == 1:
if m1 == 0 and m2 == 0:
print "adaboost-m1 or adaboost-m2 ?"
if m1 == 1 and m2 == 1:
print "adaboost-m1 or adaboost-m2 ?"
if m1 == 1 and m2 == 0:
instance = Controller()
instance.m1_controller(dirctory=config_path,
dataset=dataset,
n_y=n_y)
if m1 == 0 and m2 == 1:
instance = Controller()
instance.m2_controller(dirctory=config_path,
dataset=dataset,
n_y=n_y)
sys.exit()
except Exception as err:
print err
def plot_select():
ind = list1.curselection()
if list1.curselection() != ():
if ind[0] == 0:
sam = Samples(tipo=0)
samples = sam.get_sin(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 1:
sam = Samples(tipo=1)
samples = sam.get_square(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 2:
sam = Samples(tipo=2)
samples = sam.get_blur(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 3:
sam = Samples(tipo=3)
samples = sam.get_cuadratic(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 4:
sam = Samples(tipo=4)
samples = sam.get_diagonal_line(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 5:
sam = Samples(tipo=5)
samples = sam.get_horizontal_line(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 6:
sam = Samples(tipo=6)
samples = sam.get_vertical_line(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 7:
sam = Samples(tipo=7)
samples = sam.get_x(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 8:
sam = Samples(tipo=8)
samples = sam.get_circle(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 9:
sam = Samples(tipo=9)
samples = sam.get_curve_x(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 10:
sam = Samples(tipo=10)
samples = sam.get_diagonal_line2(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 11:
sam = Samples(tipo=11)
samples = sam.get_dependent(numberSam.get())
case_function(samples, sam)
elif ind[0] == 12:
sam = Samples(tipo=12)
samples = sam.get_independent(numberSam.get())
case_function(samples, sam)
elif ind[0] == 13:
sam = Samples(tipo=13)
samples = sam.get_corr(numberSam.get(), noise.get())
case_function(samples, sam)
elif ind[0] == 14:
sam = Samples(tipo=14)
samples = sam.get_file()
case_function(samples, sam)
else:
mitexto.set('Error')
else:
mitexto.set('Select a type')
DIR: 'data/facedata',
HEIGHT: 68,
WIDTH: 61,
LABEL: 2,
PIXELS: None
},
DIGIT: {
DIR: 'data/digitdata',
HEIGHT: 20,
WIDTH: 29,
LABEL: 10,
PIXELS: None
}
}
samples = Samples(map.get(inp).get(DIR))
dataClassifier = DataClassifier(
map.get(inp).get(HEIGHT),
map.get(inp).get(WIDTH),
map.get(inp).get(LABEL),
map.get(inp).get(PIXELS))
perceptronClassifier = PerceptronClassifier(dataClassifier.FEATURES,
dataClassifier.LABELS)
samples.readFiles()
dataset = 0
featureValueListForAllTrainingImages, actualLabelForTrainingList = dataClassifier.extractFeatures(
samples.train_lines_itr, samples.train_labelsLines_itr)
TOTALDATASET = len(featureValueListForAllTrainingImages)
HEIGHT: 68,
WIDTH: 61,
LABEL: 2,
PIXELS: None
},
DIGIT: {
DIR: 'data/digitdata',
HEIGHT: 20,
WIDTH: 29,
LABEL: 10,
PIXELS: None
}
}
dataType = map.get(inp)
samples = Samples(dataType.get(DIR))
dataClassifier = DataClassifier(dataType.get(HEIGHT), dataType.get(WIDTH),
dataType.get(LABEL), dataType.get(PIXELS),
gridSize)
samples.readFiles()
''' Extracting Features from the Training Data '''
featureValueListForAllTrainingImages, actualLabelForTrainingList = \
dataClassifier.extractFeatures(samples.train_lines_itr, samples.train_labelsLines_itr)
runErrorList = []
runTimeList = []
runDataSetIncrements = []
for _ in range(0, 3):
