I want TPP to be an array with the TPP value for each threshold. The print should be like: TPP is: n1, n2...
threshold=[0, 6, 15]
df=[3,13,19,21]
y_true=[0,0,1,0]
y_pred=np.where(df<threshold,1,0)
cm=confusion_matrix(y_true,y_pred)
TP=cm[0,0]
FN=cm[0,1]
TPP=TP/(TP+FN)
print('TPP is:',TPP)
It seems to me that this code achieves your goal:
from sklearn.metrics import confusion_matrix
import numpy as np
threshold = [0, 6, 15, 20]
df = [3,13,19,21]
y_true = [0,0,1,0]
print('TPP is:', end=' ')
for idx, i in enumerate(threshold):
y_pred = np.where(np.array(df) < i,1,0)
are_ones = y_true == np.ones(len(y_true))
predicted_as_zero = y_pred == np.zeros(len(y_true))
TP = sum( are_ones & np.array(y_pred) )
FN = sum( are_ones & predicted_as_zero )
if idx+1 == len(threshold):
print(TP/(TP+FN), end='')
else:
print(TP/(TP+FN), end=', ')