1 使用numpy计算true positives等
import numpy as np
y_true = np.array([0, 1, 1, 0, 1, 0])
y_pred = np.array([1, 1, 1, 0, 0, 1])
# true positive
TP = np.sum(np.multiply(y_true, y_pred))
print(TP)
# false positive
FP = np.sum(np.logical_and(np.equal(y_true, 0), np.equal(y_pred, 1)))
print(FP)
# false negative
FN = np.sum(np.logical_and(np.equal(y_true, 1), np.equal(y_pred, 0)))
print(FN)
# true negative
TN = np.sum(np.logical_and(np.equal(y_true, 0), np.equal(y_pred, 0)))
print(TN)
输出结果：
2
2
1
1
2 使用tensorflow计算true positives等
import tensorflow as tf
sess = tf.Session()
y_true = tf.constant([0, 1, 1, 0, 1, 0])
y_pred = tf.constant([1, 1, 1, 0, 0, 1])
# true positive
TP = tf.reduce_sum(tf.multiply(y_true, y_pred))
print(sess.run(TP))
# false positive
FP = tf.reduce_sum(tf.cast(tf.logical_and(tf.equal(y_true, 0), tf.equal(y_pred, 1)), tf.int32))
print(sess.run(FP))
# false negative
FN = tf.reduce_sum(tf.cast(tf.logical_and(tf.equal(y_true, 1), tf.equal(y_pred, 0)), tf.int32))
print(sess.run(FN))
# true negative
TN = tf.reduce_sum(tf.cast(tf.logical_and(tf.equal(y_true, 0), tf.equal(y_pred, 0)), tf.int32))
print(sess.run(TN))
输出结果：
2
2
1
1
3 使用sklearn的metrics模块计算precision，recall和f1-score
3.1 数据是list类型
from sklearn.metrics import precision_score, recall_score, f1_score
y_true = [0, 1, 1, 0, 1, 0]
y_pred = [1, 1, 1, 0, 0, 1]
p = precision_score(y_true, y_pred, average='binary')
r = recall_score(y_true, y_pred, average='binary')
f1score = f1_score(y_true, y_pred, average='binary')
print(p)
print(r)
print(f1score)
输出结果：
0.5
0.666666666667
0.571428571429
3.2 数据是ndarray类型
from sklearn.metrics import precision_score, recall_score, f1_score
import numpy as np
y_true = np.array([[0, 1, 1],
[0, 1, 0]])
y_pred = np.array([[1, 1, 1],
[0, 0, 1]])
y_true = np.reshape(y_true, [-1])
y_pred = np.reshape(y_pred, [-1])
p = precision_score(y_true, y_pred, average='binary')
r = recall_score(y_true, y_pred, average='binary')
f1score = f1_score(y_true, y_pred, average='binary')
print(p)
print(r)
print(f1score)
输出结果：
0.5
0.666666666667
0.571428571429
