def CreateFigures(self, params, fitOrSim):
if fitOrSim == 'fit':
graphLabel = 'Full complex fit: '
elif fitOrSim == 'sim':
graphLabel = 'Simulated using: '
else:
graphLabel = ''
plt.close('all')
# make layout for graphs
gs0 = gridspec.GridSpec(1, 2)
gs00 = gridspec.GridSpecFromSubplotSpec(4, 3, subplot_spec=gs0[0])
gs01 = gridspec.GridSpecFromSubplotSpec(4, 4, subplot_spec=gs0[1])
f = plt.figure(1, figsize=[8, 3.5], tight_layout='true')
###########################################################
#### PLOT NYQUIST COMBINED FITTINGS ####
###########################################################
nyGraph = plt.Subplot(f, gs01[:, :])
f.add_subplot(nyGraph)
nyGraph.plot(self.activeData.zPrime,
self.activeData.ZdoublePrime,
'bo',
ls='--',
markersize=2,
linewidth=1,
label='data: ' + self.activeData.dataNameNoExt)
nyGraph.plot(
self.funcreal(params),
self.funcImg(params),
'ro',
markersize=2,
label=
'\n%s\nLwire=%.5e\nRmem=%5.8f\nRcl=%5.8f\nQdl=%5.5f\nphi=%5.5f' %
(graphLabel, params[0] * float(self.area.IE.get()),
params[1] * float(self.area.IE.get()),
params[2] * float(self.area.IE.get()), params[3], params[4]))
plt.gca().invert_yaxis()
plt.xticks(rotation=20)
plt.xlabel('Re(Z)')
plt.ylabel('Im(Z)')
plt.legend(loc=2, fontsize=6)
###########################################################
#### PLOT Phase vs w COMBINED FITTINGS ####
###########################################################
phaseGraph = plt.Subplot(f, gs00[-4, :3])
f.add_subplot(phaseGraph)
phaseGraph.plot(self.activeData.frequency,
self.activeData.phase,
'bo',
ls='--',
markersize=2,
linewidth=1)
phaseGraph.plot(self.activeData.frequency,
self.funcPhase(params),
'ro',
markersize=2)
plt.ylabel('phase')
#plt.gca().set_yscale('log')
plt.gca().invert_yaxis()
plt.gca().set_xscale('log')
###########################################################
#### PLOT |Z| vs w COMBINED FITTINGS ####
###########################################################
modZgraph = plt.Subplot(f, gs00[-3, :3])
f.add_subplot(modZgraph)
modZgraph.plot(self.activeData.frequency,
self.activeData.modZExperimentalComplex,
'bo',
ls='--',
markersize=2,
linewidth=1)
modZgraph.plot(self.activeData.frequency,
self.funcAbs(params),
'ro',
markersize=2)
plt.ylabel('|Z|')
plt.gca().set_yscale('log')
plt.gca().set_xscale('log')
###########################################################
#### PLOT Im(Z) ####
###########################################################
imZgraph = plt.Subplot(f, gs00[-2, :3])
f.add_subplot(imZgraph)
imZgraph.plot(self.activeData.frequency,
self.activeData.ZdoublePrime,
'bo',
ls='--',
markersize=2,
linewidth=1)
imZgraph.plot(self.activeData.frequency, (self.funcImg(params)),
'ro',
markersize=2)
plt.ylabel('Im(Z)')
plt.gca().set_yscale('linear')
plt.gca().set_xscale('log')
plt.gca().set_xticks([])
###########################################################
#### PLOT Re(Z) ####
###########################################################
reZgraph = plt.Subplot(f, gs00[-1, :3])
f.add_subplot(reZgraph)
reZgraph.plot(self.activeData.frequency,
self.activeData.zPrime,
'bo',
ls='--',
markersize=2,
linewidth=1)
reZgraph.plot(self.activeData.frequency,
self.funcreal(params),
'ro',
markersize=2)
plt.xlabel('frequency')
plt.ylabel('Re(Z)')
plt.gca().set_yscale('log')
plt.gca().set_xscale('log')
###########################################################
#### Draw figure in Tkinter ####
###########################################################
for widget in self.plotFrame.winfo_children():
widget.destroy()
for widget in self.plotFrameToolBar.winfo_children():
widget.destroy()
dataPlot = FigureCanvasTkAgg(f, master=self.plotFrame)
dataPlot.close_event()
dataPlot.draw()
dataPlot.get_tk_widget().grid(
row=0,
sticky=N + S + E + W,
)
toolbar = NavigationToolbar2Tk(dataPlot, self.plotFrameToolBar)
toolbar.update()
dataPlot._tkcanvas.grid(row=0, sticky=W + S)
self.plotToSave = f
print('done with plotting')
def show_entry_fields():
global dataset, dataset_test, labels, labels_test, labels_names, probCLF, features_names, labels_test, labels_test_result
dataset,dataset_test,labels,labels_test,features_names = functions_RP.inicialize_data()
try:
dataset, dataset_test,labels_test,labels,features_names
except NameError:
label_classificar.configure(text="Não pode Classificar. Tem de fazer load de tudo!")
else:
dataset1 = dataset
dataset_test1= dataset_test
labels1= labels
labels_test1= labels_test
features_names1 = features_names
if(var.get() == 0 or var3.get() == 0):
print("Nao pode classificar")
label_classificar.configure(text="Dados insuficientes para classificar! ")
else:
label_classificar.configure(text="")
if(var4.get() == 1):
dataset1, dataset_test1 = functions_RP.pre_processamento(dataset1, dataset_test1)
if(var.get() != 0):
labels1, labels_test1, labels_names1, probCLF = functions_RP.define_labels(var.get())
if(var1.get() != 0):
features_names1, dataset1, dataset_test1 = functions_RP.features_selection(dataset1, dataset_test1, features_names1, labels1, var1.get(), 30, probCLF)
if(var2.get() != 0):
dataset1, dataset_test1 = functions_RP.features_reduction(dataset1, dataset_test1, labels1, var2.get(), 10)
if(var3.get() == 1):
labels_test_result = functions_RP.distMinClf_CV(dataset1, labels1, dataset_test1, labels_test1, probCLF, 'euclidean', labels_names)
elif (var3.get() == 2):
labels_test_result = functions_RP.distMinClf_CV(dataset1, labels1, dataset_test1, labels_test1, probCLF, 'mahalanobis', labels_names)
elif(var3.get() == 3):
labels_test_result = functions_RP.clfNaiveBayes(dataset1, dataset_test1, labels1, labels_test1, labels_names, probCLF)
elif (var3.get() == 4):
labels_test_result = functions_RP.clfSVM(dataset1, dataset_test1, labels1, labels_test1, labels_names, probCLF)
elif(var3.get() == 5):
labels_test_result = functions_RP.clfKNN(dataset1, dataset_test1, labels1, labels_test1, labels_names, probCLF)
elif(var3.get() == 6):
labels_test_result = functions_RP.clfRandomForest(dataset1, dataset_test1, labels1, labels_test1, labels_names, probCLF)
elif (var3.get() == 7):
labels_test_result = functions_RP.clfQDA(dataset1, dataset_test1, labels1, labels_test1, labels_names, probCLF)
elif (var3.get() == 8):
labels_test_result = functions_RP.clfLDA(dataset1, dataset_test1, labels1, labels_test1, labels_names, probCLF)
if(probCLF == 'bin'):
report, acc, matrix, rocScore = functions_RP.metricsClf(labels_test_result, labels_test1, labels_names, probCLF)
label_report.configure(text=report)
text_acc = "Accuracy: " + str(acc)
label_aac.configure(text=text_acc)
label_aac.grid_configure(row=7, column=2, columnspan=5)
text_roc = "AUC: " + str(rocScore)
label_roc_auc.configure(text=text_roc)
text_matrix = "Matriz de Confusão: \n 0 1 \n0 "+ str(matrix[0,0]) + " " + str(matrix[0,1]) + "\n1 " + str(matrix[1,0]) + " " + str(matrix[1,1])
label_matrix.configure(text=text_matrix)
label_matrix.grid_configure(row=9, column=2, columnspan=5, rowspan=4, sticky=N)
frp, tpr, roc_auc = functions_RP.ROC(1, labels_test1, labels_test_result)
f.set_visible(True)
f.clear()
f.add_axes
canvas = FigureCanvasTkAgg(f)
lw = 2
a = f.add_subplot(111)
a.set_xlabel('True Positive Rate')
a.set_ylabel('True Positive Rate')
a.set_title('Receiver operating characteristic')
a.axis((0.0,1.0,0.0,1.0))
a.plot(frp, tpr, color='darkorange',lw=lw, label='ROC curve (area = %0.2f)' % roc_auc)
canvas.get_tk_widget().grid(row=17, column=2, rowspan=10, columnspan=6, padx=50, sticky=W, pady=20)
else:
report, acc, matrix = functions_RP.metricsClf(labels_test_result, labels_test1, labels_names, probCLF)
f.clear()
f.set_visible(False)
canvas = FigureCanvasTkAgg(f)
canvas.get_tk_widget().grid(row=17, column=2, rowspan=10, columnspan=6, padx=50, sticky=W, pady=20)
canvas.close_event()
report, acc, matrix = functions_RP.metricsClf(labels_test_result, labels_test1, labels_names, probCLF)
label_report.configure(text=report)
label_report.grid_configure(row=1, column=2, rowspan=9, columnspan=5, sticky=N)
text_acc = "Accuracy: " + str(acc)
label_aac.configure(text=text_acc)
label_roc_auc.configure(text="")
label_aac.grid_configure(row=10, column=2, columnspan = 5, sticky=N)
text_matrix = "Matriz de Confusão: \n 1 2 3 4 5 6"
for i in range(1,7):
text_matrix = text_matrix + "\n" + str(i) + " "
for j in range(0,6):
text_matrix= text_matrix + str(matrix[i-1,j]) + " "
print(text_matrix)
label_matrix.configure(text=text_matrix)
label_matrix.grid_configure(row=11, column=2, columnspan = 6, rowspan=8, sticky=N)
