def __init__(self, *args, **kwargs):
QWidget.__init__(self)
self.fcList = db.getFlowcytometerList()
self.setWindowIcon(QIcon('logo.png'))
self.setWindowTitle('Flow cytometer configuration')
self.resize(300, 300)
self.move(500, 200)
self.label = QLabel('Select the active flow cytometer')
self.label.setFont(QFont('Arial', 13, QFont.Bold))
db.updateFc()
self.labelClass = QLabel('\nConfigure a flow cytometer')
self.aClassQ = QComboBox()
for a in self.fcList:
self.aClassQ.addItem(a)
# todo del the fc name if id not in channels column fcid
v = db.getFcValue()
if v != []:
self.aClassQ.setCurrentIndex(self.fcList.index(v))
self.buttonAdd = QPushButton('Add a flow cytometer')
self.buttonAdd.clicked.connect(self.addClass)
self.buttonDel = QPushButton('Delete a flow cytometer')
self.buttonDel.clicked.connect(self.delClass)
self.ok = QPushButton('Update a flow cytometer')
self.ok.clicked.connect(self.updateFc)
self.ok1 = QPushButton('OK')
self.ok1.clicked.connect(self.validate)
self.closeB = QPushButton('Exit')
self.closeB.clicked.connect(self.closeW)
layout = QGridLayout()
layout.addWidget(self.label, 0, 0, 1, 2)
layout.addWidget(self.aClassQ, 1, 0, 1, 2)
layout.addWidget(self.labelClass, 2, 0)
layout.addWidget(self.ok, 4, 0, 1, 2)
layout.addWidget(self.buttonAdd, 3, 0, 1, 2)
layout.addWidget(self.buttonDel, 5, 0, 1, 2)
layout.addWidget(self.closeB, 6, 1)
layout.addWidget(self.ok1, 6, 0)
self.setLayout(layout)
self.show()
def __init__(self, state):
QWidget.__init__(self)
v = db.getFcValue()
if v == []:
msg = QMessageBox.warning(
self, 'Error',
'Please select a flow cytometer to launch the program')
self.show()
self.close()
return
self.setWindowIcon(QIcon('logo.png'))
self.state = state
if self.state == 'clustering':
self.setWindowTitle('Clustering step 1/4')
self.title = QLabel(
'Cluster prediction of community\n \n') #TODO
elif self.state == 'assessment':
self.setWindowTitle('Predict step 1/4')
self.title = QLabel(
'Species prediction\nfor in-silico community\n \n')
else:
self.setWindowTitle('Predict step 1/4')
self.title = QLabel('Species prediction\nin community\n \n')
self.resize(400, 100)
self.move(500, 200)
self.title.setFont(QFont('Arial', 13, QFont.Bold))
self.title.setAlignment(Qt.AlignCenter)
classList = db.getClassNames(True)
self.classes = QComboBox()
for a in classList:
self.classes.addItem(a)
self.ok1 = QPushButton('OK')
self.ok1.clicked.connect(self.classUpdate)
self.grid = QGridLayout()
self.grid.addWidget(self.title, 0, 0, 1, 2)
self.labelClass = QLabel('Class ')
self.grid.addWidget(self.labelClass, 2, 0)
self.grid.addWidget(self.classes, 2, 1)
self.grid.addWidget(self.ok1, 3, 1)
self.setLayout(self.grid)
def validate(self):
db.desactivateFc()
db.saveFc(self.fcList[self.aClassQ.currentIndex()])
db.removeUnusedParam()
if db.getFcValue() != []:
self.aClassQ.setCurrentIndex(self.fcList.index(db.getFcValue()))
def __init__(self, *args, **kwargs):
QWidget.__init__(self)
v = db.getFcValue()
if v == []:
msg = QMessageBox.warning(
self, 'Error',
'Please select a flow cytometer to launch the program')
self.show()
self.close()
return
self.setWindowIcon(QIcon('logo.png'))
self.layout = QHBoxLayout(self)
self.scrollArea = QScrollArea(self)
self.scrollArea.setWidgetResizable(True)
self.scrollAreaWidgetContents = QWidget()
self.grid = QGridLayout(self.scrollAreaWidgetContents)
self.scrollArea.setWidget(self.scrollAreaWidgetContents)
self.layout.addWidget(self.scrollArea)
title = QLabel('Advanced parameters')
title.setAlignment(Qt.AlignCenter)
title.setFont(QFont("Century Gothic", 15, QFont.Bold))
self.fig = ['Save', 'Show', 'None']
self.gat = ['None', 'Line', 'Machine']
self.met = [
'Neural network', 'Random forest', 'Logistic regression',
'Random guess'
]
#self.graphOption = ['FSC-A', 'SSC-A', 'FL1-A', 'FL2-A', 'FL3-A', 'FL4-A', 'FSC-H', 'SSC-H', 'FL1-H', 'FL2-H',
# 'FL3-H', 'FL4-H', 'Width']
channels, self.graphOption, dic = db.getChannels()
repeatL = QLabel('Number of runs ')
self.repeat = QSpinBox(self)
self.repeat.setRange(1, 50)
averageL = QLabel('Create average prediction from runs')
self.average = QCheckBox()
doubtL = QLabel(
'Percentage minimum of appearance \nto validate a prediction')
self.doubt = QSpinBox(self)
self.doubt.setRange(0, 100)
nbCl = QLabel('Training:\n Number of cells per record')
self.nbC = QSpinBox(self)
self.nbC.setRange(100, 10000)
nbC2l = QLabel('Tool analysis:\n Number of cells per record')
self.nbC2 = QSpinBox(self)
self.nbC2.setRange(0, 10000)
ratioL = QLabel('Ratio learning/test')
self.ratio = QDoubleSpinBox()
self.ratio.setRange(0, 2)
self.ratio.setStepType(QAbstractSpinBox.AdaptiveDecimalStepType)
figureL = QLabel('Figures view mode')
self.figure = QComboBox()
self.figure.addItems(self.fig)
gatingL = QLabel('Gating type ')
self.gating = QComboBox()
self.gating.addItems(self.gat)
self.ok = QPushButton('Save Settings')
self.ok.clicked.connect(self.saveParam)
self.default = QPushButton('Back on default')
self.default.clicked.connect(partial(self.maj, True))
self.methodL = QLabel("Machine learning method")
self.method = QComboBox()
self.method.addItems(self.met)
self.showGat = QCheckBox()
self.showGatL = QLabel('Show gating effect')
self.clustDist = QDoubleSpinBox(self)
self.clustDist.setRange(0, 5)
self.clustDist.setStepType(QAbstractSpinBox.AdaptiveDecimalStepType)
self.clustDistL = QLabel('Cluster distance %')
self.axesL = QLabel('Select 3D graph axes')
self.axe1 = QComboBox()
self.axe1.addItems(self.graphOption)
self.axe2 = QComboBox()
self.axe2.addItems(self.graphOption)
self.axe3 = QComboBox()
self.axe3.addItems(self.graphOption)
self.grid.addWidget(title, 0, 0, 1, 4)
self.grid.addWidget(repeatL, 1, 0)
self.grid.addWidget(self.repeat, 1, 1)
self.grid.addWidget(averageL, 2, 0)
self.grid.addWidget(self.average, 2, 1)
self.grid.addWidget(doubtL, 3, 0)
self.grid.addWidget(self.doubt, 3, 1)
self.grid.addWidget(figureL, 4, 0)
self.grid.addWidget(self.figure, 4, 1)
self.grid.addWidget(gatingL, 1, 2)
self.grid.addWidget(self.gating, 1, 3)
self.grid.addWidget(nbCl, 3, 2)
self.grid.addWidget(self.nbC, 3, 3)
self.grid.addWidget(nbC2l, 4, 2)
self.grid.addWidget(self.nbC2, 4, 3)
self.grid.addWidget(self.methodL, 5, 2)
self.grid.addWidget(self.method, 5, 3)
self.grid.addWidget(ratioL, 6, 2)
self.grid.addWidget(self.ratio, 6, 3)
self.grid.addWidget(self.showGatL, 2, 2)
self.grid.addWidget(self.showGat, 2, 3)
self.grid.addWidget(self.axesL, 5, 0)
self.grid.addWidget(self.axe1, 5, 1)
self.grid.addWidget(self.axe2, 6, 1)
self.grid.addWidget(self.axe3, 7, 1)
self.grid.addWidget(self.clustDistL, 8, 0)
self.grid.addWidget(self.clustDist, 8, 1)
self.grid.addWidget(self.ok, 9, 0, 2, 2)
self.grid.addWidget(self.default, 9, 2, 2, 2)
self.setWindowTitle("Settings")
self.resize(900, 500)
self.move(50, 200)
self.maj(False)
def run(self):
"""
Function from the interface launching the machine learning program
:param self:
:return:
"""
channels, replaced, dicChannels = db.getChannels()
# prediction ou assessment
typeA = self.state
# Species
species = []
sp = self.selectedSpecies
# reference file from the database
files = []
for s in sp:
ref = db.getReferences(s, self.parent.aClass)
for i in range(len(ref)):
files = files + ['references/' + ref[i]]
species = species + [s]
# Analysis files:
files2 = self.files
# Parameter from the database : gating number repeat...
repeat = int(db.getParamValue('reapt'))
param = [
db.getParamValue('graph1'),
db.getParamValue('graph2'),
db.getParamValue('graph3')
]
doubt = int(db.getParamValue('doubt')) / 100
fc = db.getFcValue()
nbC = int(db.getParamValue('nbC'))
ratio = float(db.getParamValue('ratio'))
var = float(db.getParamValue('clustDist'))
average = db.getParamValue('average')
if average in ['True', '1', 1, True]:
average = True
else:
average = False
figure = db.getParamValue('figure')
if figure == 'None':
figure = None
elif figure == 'Save':
figure = 'save'
else:
figure = 'show'
gating = db.getParamValue('gating')
if gating == 'None' or gating is None or gating == 0:
gating = None
elif gating == 'Line':
gating = 'line'
else:
gating = 'machine'
method = db.getParamValue('method')
if method == 'Neural network':
method = 'neur'
elif method == 'Random forest':
method = 'rand'
elif method == 'Logistic regression':
method = 'log'
elif method == 'Random guess':
method = 'rdguess'
showGating = db.getParamValue('showGat')
if showGating in [True, 1, 'True', '1']:
showgat = True
else:
showgat = False
nbC2 = int(db.getParamValue('nbC2'))
if nbC2 == 0:
nbC2 = None
if typeA == 'prediction':
files2 = files2[0]
species2 = ['unknown'] * len(files2)
directory = predictions(files,
species,
files2,
species2,
nbC,
None,
gating=gating,
showgat=showgat,
predAn='prediction',
predtype=method,
ratio=ratio,
repeat=repeat,
average=average,
doubt=doubt,
random_state=0,
save=figure,
fc=fc,
param=param,
channels=replaced,
dicChannels=dicChannels)
elif typeA == 'assessment': # case of assessment
species2 = []
nf = []
for i in range(len(files2)):
nf = nf + files2[i]
species2 = species2 + ([sp[i]] * len(files2[i]))
files2 = nf.copy()
directory = predictions(files,
species,
files2,
species2,
nbC,
nbC2,
gating=gating,
showgat=showgat,
predAn='analysis',
predtype=method,
ratio=ratio,
repeat=repeat,
average=average,
doubt=doubt,
random_state=0,
save=figure,
fc=fc,
param=param,
channels=replaced,
dicChannels=dicChannels)
elif typeA == 'clustering':
files2 = files2[0]
species2 = ['unknown'] * len(files2)
directory = c.clustering(files,
species,
files2,
species2,
predAn='prediction',
param=param,
nbC=nbC,
nbC2=None,
save=figure,
var=var,
gating=gating,
showgat=showgat,
fc=fc,
method=2,
channels=replaced,
dicChannels=dicChannels)
elif typeA == 'clustA':
species2 = []
nf = []
for i in range(len(files2)):
nf = nf + files2[i]
species2 = species2 + ([sp[i]] * len(files2[i]))
files2 = nf.copy()
directory = c.clustering(files,
species,
files2,
species2,
predAn='analysis',
param=param,
nbC=nbC,
nbC2=nbC2,
save=figure,
var=var,
method=2,
gating=gating,
showgat=showgat,
fc=fc,
channels=replaced,
dicChannels=dicChannels)
else:
directory = 'None'
return directory