Python NumPy中 -1 的作用

倒数第一

【TODO】
在list, tuple, array中表示倒数第一个
简单举例

a01 = [3, 2]                                        
print("a01[:-1]:", a01[:-1])        # output: 3     
print("a01[0:-1]:", a01[0:-1])      # output: 3     

大原则：左闭（inclusive）右开（exclusive）原则
np.random.randint
randint(low, high=None, size=None, dtype=‘l’)
Return random integers from low (inclusive) to high (exclusive).
Return random integers from the “discrete uniform” distribution of
the specified dtype in the “half-open” interval [low, high). If
high is None (the default), then results are from [0, low).
返回值：[low, high)
复杂举例
这段函数是从keras.utils.to_categorical复制过来的，将其改名为my_to_categorical，作用是生成one-hot码。

#! /usr/bin/env python
# -*- coding: utf-8 -*-

# import the necessary packages
import numpy as np

def my_to_categorical(y, num_classes=None, dtype='float32'):
    y = np.array(y, dtype='int')
    input_shape = y.shape
    if input_shape and input_shape[-1] == 1 and len(input_shape) > 1:
        input_shape = tuple(input_shape[:-1])
    y = y.ravel()
    if not num_classes:
        num_classes = np.max(y) + 1
    n = y.shape[0]
    categorical = np.zeros((n, num_classes), dtype=dtype)
    categorical[np.arange(n), y] = 1
    output_shape = input_shape + (num_classes,)
    categorical = np.reshape(categorical, output_shape)
    return categorical

a1 = np.random.randint(10, size=(3, 1))     # a1.shape = (3,1)
print(a1.shape)
print(a1)
print("#"*6, "my_to_categorical", "#"*6)
newA1 = my_to_categorical(a1, num_classes=10)
print(newA1)    

自动推断

通过已知参数推断出的一个形状参数时，可以将其设置为-1.
【注意】不可以写成a2 = a1.reshape(2, -1, -1)
会报错：“can only specify one unknown dimension”只能指定一个未知维度

# import the necessary packages
import numpy as np

a1 = np.arange(36).reshape(4,9)     # a1.shape = (4, 9)
a2 = a1.reshape(2, 3, -1)       # 6 = 36 / (2*3)
print("a2.shape：",a2.shape)     # a2.shape = (2， 3， 6)
a3 = a1.reshape(2, -1, 6)       # 3 = 36 / (2*6)
print("a3.shape：",a3.shape)     # a3.shape = (2， 3， 6)