In [38]:
import matplotlib.pyplot as plt
import numpy as np
# Create a color array based on TP, FP, TN, FN
colors = []
for i in range(len(label_curve)):
label = label_curve[i]
pred = predicted_curve[i]
if label > 0 and pred >= 0.5: # TP: Actual positive and predicted as positive
colors.append('green')
elif label == 0 and pred < 0.5: # TN: Actual negative and predicted as negative
colors.append('blue')
elif label == 0 and pred >= 0.5: # FP: Actual negative but predicted as positive
colors.append('red')
else: # FN: Actual positive but predicted as negative
colors.append('orange')
# Convert colors to a numpy array
colors = np.array(colors)
# Create a scatter plot of main_curve with color-coding based on prediction correctness
plt.scatter(range(len(main_curve)), main_curve, c=colors, s=15) # s parameter controls the size of dots
plt.plot(range(len(main_curve)), main_curve, color='black', linewidth=0.5) # plot the line
# Title and labels
plt.title('Example Signal with Prediction Outcomes')
plt.xlabel('Time Step (index)')
plt.ylabel('Example Signal')
# Setting the x limit
plt.xlim([400, 600])
# Legend
plt.scatter([], [], color='green', label='True Positive', s=15)
plt.scatter([], [], color='blue', label='True Negative', s=15)
plt.scatter([], [], color='red', label='False Positive', s=15)
plt.scatter([], [], color='orange', label='False Negative', s=15)
plt.legend()
plt.savefig('Example_plot_'+station+'_'+sign+'.png')
plt.show()
