np.around、np.reshape、np.squeeze
1、np.around
1、四舍五入
2、当是.5时,最后取得证书要是偶数
In [17]: np.around([0.4,0.5,0.12,0.6,0.3,1.2,1.8,1.3])
Out[17]: array([0., 0., 0., 1., 0., 1., 2., 1.])In [21]: np.round([0.5,1.5,2.5,3.5,4.5,5.5,6.5])
Out[21]: array([0., 2., 2., 4., 4., 6., 6.])
2、np.reshape
体会-1的意义
1、Reshape (num_px, num_px, 3) ---> (num_px ∗num_px ∗ 3, 1).
一个简单的小技巧: shape (a,b,c,d) ---> shape (b∗c∗d, a)
X_flatten = X.reshape(X.shape[0], -1).T # X.T is the transpose of X
2、reshape(1,-1) 返回行向量
In [47]: a.reshape(1,-1)
Out[47]: array([[1, 2, 3, 4, 5, 6, 7, 8]])
3、reshape(-1,1)返回列向量
In [48]: a.reshape(-1,1)
Out[48]:array([[1],[2],[3],[4],[5], [6],[7],[8]])
3、np.squeeze
去掉shape为1的维度
cost = np.squeeze(cost) # makes sure cost is the dimension we expect.
# E.g., turns [[17]] into 17