0
Grafiği görüntülediğimde, ek açıklama açıkça görülebilir, ancak bunu kaydetmeye çalıştığımda, bunu görüyorum (bkz. Altta, ek açıklama görülebilir):Matplotlib - çizim kaydedildiğinde, ek açıklama gösterilmiyor
Kaydedilecek alanı nasıl arttıracağınız bir yolu var mı? Aşağıda
Kodum:
def plot_conMatrix_ROC(pred_y, test_y, notes, save=False):
"""Arguments:
pred_y: my prediction of data
test_y: true values to compare with my prediction
notes: optional, if i want add any notes to the plot
save: option to save the plot
"""
"""------------- Confusion matrix ------------------"""
#Calculate metrics
precision = precision_score(test_y,pred_y)
recall = recall_score(test_y,pred_y)
f1 = f1_score(test_y,pred_y)
accuracy = accuracy_score(test_y, pred_y)
#calculate confusion matrix
cm = confusion_matrix(test_y,pred_y)
class_names = np.unique(test_y)
#normalize to scale for the coloring
norm_conf = []
for i in cm:
a = 0
tmp_arr = []
a = sum(i, 0)
for j in i:
tmp_arr.append(float(j)/float(a))
norm_conf.append(tmp_arr)
#plot the matrix
fig = plt.figure('conMat')
axes = fig.add_subplot(121)
fig.set_figheight(5)
fig.set_figwidth(9)
res=axes.imshow(np.array(norm_conf),
cmap=plt.cm.Purples,interpolation='nearest')
width = len(norm_conf)
height = len(norm_conf[0])
#insert numbers in the matrix
for x in xrange(width):
for y in xrange(height):
axes.annotate(str(cm[x][y]), xy=(y, x),
horizontalalignment='center',
verticalalignment='center')
#Describe the chart
plt.xticks(range(width), class_names)
plt.yticks(range(height), class_names)
plt.ylabel('True label')
plt.xlabel('Predicted label')
plt.title('Confusion matrix')
txt = 'Accuracy: %0.2f, Recall: %0.2f, Precision: %0.2f, F1: %0.2f' % (accuracy,recall,precision,f1)
#Annotate
axes.annotate(txt, xy=(1, 0), xycoords='axes fraction', fontsize=13,
xytext=(60, -40), textcoords='offset points',
ha='center', va='center')
if notes <> '':
axes.annotate(notes, xy=(1, 0), xycoords='axes fraction', fontsize=13, xytext=(90, -70), textcoords='offset,ha='center,va='center')
"""------------- ROC ------------------"""
#calculate AUC
fpr, tpr, _ = roc_curve(test_y, pred_y)
roc_auc = np.trapz(fpr,tpr)
axes = fig.add_subplot(122)
fig.set_figheight(4)
fig.set_figwidth(9)
plt.plot(fpr,tpr,'b',label='AUC = %0.2f'% roc_auc)
plt.plot([0,1],[0,1],linestyle='--',color=(0.6, 0.6, 0.6), label='random guessing')
plt.xlabel('False positive rate')
plt.ylabel('True positive rate')
plt.title('ROC curve')
plt.legend(loc='lower right')
if save:
plt.savefig('CM_ROC.png')
fig.tight_layout()
plt.show()
plt.clf()
Ve çağrı:
plot_conMatrix_ROC(pred, y_test, notes='test 2',save=True)