极简代码（一）—— 精确率和错误率的计算

numpy

# precision
np.mean(y_pred == y_true)

# error
1 - np.mean(y_pred == y_true)

theano

# error
T.mean(T.neg(y_pred, y_true))

# precision
1 - T.mean(T.neg(y_pred, y_true))