mirror of
https://github.com/apachecn/ailearning.git
synced 2026-02-03 18:34:17 +08:00
241 lines
2.7 KiB
Markdown
241 lines
2.7 KiB
Markdown
# 二元运算
|
||
|
||
In [1]:
|
||
|
||
```py
|
||
import numpy as np
|
||
|
||
```
|
||
|
||
## 四则运算
|
||
|
||
| 运算 | 函数 |
|
||
| --- | --- |
|
||
| `a + b` | `add(a,b)` |
|
||
| `a - b` | `subtract(a,b)` |
|
||
| `a * b` | `multiply(a,b)` |
|
||
| `a / b` | `divide(a,b)` |
|
||
| `a ** b` | `power(a,b)` |
|
||
| `a % b` | `remainder(a,b)` |
|
||
|
||
以乘法为例，数组与标量相乘，相当于数组的每个元素乘以这个标量：
|
||
|
||
In [2]:
|
||
|
||
```py
|
||
a = np.array([1,2])
|
||
a * 3
|
||
|
||
```
|
||
|
||
Out[2]:
|
||
|
||
```py
|
||
array([3, 6])
|
||
```
|
||
|
||
数组逐元素相乘：
|
||
|
||
In [3]:
|
||
|
||
```py
|
||
a = np.array([1,2])
|
||
b = np.array([3,4])
|
||
a * b
|
||
|
||
```
|
||
|
||
Out[3]:
|
||
|
||
```py
|
||
array([3, 8])
|
||
```
|
||
|
||
使用函数：
|
||
|
||
In [4]:
|
||
|
||
```py
|
||
np.multiply(a, b)
|
||
|
||
```
|
||
|
||
Out[4]:
|
||
|
||
```py
|
||
array([3, 8])
|
||
```
|
||
|
||
事实上，函数还可以接受第三个参数，表示将结果存入第三个参数中：
|
||
|
||
In [5]:
|
||
|
||
```py
|
||
np.multiply(a, b, a)
|
||
|
||
```
|
||
|
||
Out[5]:
|
||
|
||
```py
|
||
array([3, 8])
|
||
```
|
||
|
||
In [6]:
|
||
|
||
```py
|
||
a
|
||
|
||
```
|
||
|
||
Out[6]:
|
||
|
||
```py
|
||
array([3, 8])
|
||
```
|
||
|
||
## 比较和逻辑运算
|
||
|
||
| 运算 | 函数< |
|
||
| --- | --- |
|
||
| `==` | `equal` |
|
||
| `!=` | `not_equal` |
|
||
| `>` | `greater` |
|
||
| `>=` | `greater_equal` |
|
||
| `<` | `less` |
|
||
| `<=` | `less_equal` |
|
||
| | `logical_and` |
|
||
| | `logical_or` |
|
||
| | `logical_xor` |
|
||
| | `logical_not` |
|
||
| `&` | `bitwise_and` |
|
||
| | `bitwise_or` |
|
||
| `^` | `bitwise_xor` |
|
||
| `~` | `invert` |
|
||
| `>>` | `right_shift` |
|
||
| `<<` | `left_shift` |
|
||
|
||
等于操作也是逐元素比较的：
|
||
|
||
In [7]:
|
||
|
||
```py
|
||
a = np.array([[1,2,3,4],
|
||
[2,3,4,5]])
|
||
b = np.array([[1,2,5,4],
|
||
[1,3,4,5]])
|
||
a == b
|
||
|
||
```
|
||
|
||
Out[7]:
|
||
|
||
```py
|
||
array([[ True, True, False, True],
|
||
[False, True, True, True]], dtype=bool)
|
||
```
|
||
|
||
这意味着，如果我们在条件中要判断两个数组是否一样时，不能直接使用
|
||
|
||
```py
|
||
if a == b:
|
||
```
|
||
|
||
而要使用：
|
||
|
||
```py
|
||
if all(a==b):
|
||
```
|
||
|
||
对于浮点数，由于存在精度问题，使用函数 `allclose` 会更好：
|
||
|
||
```py
|
||
if allclose(a,b):
|
||
```
|
||
|
||
`logical_and` 也是逐元素的 `and` 操作：
|
||
|
||
In [8]:
|
||
|
||
```py
|
||
a = np.array([0,1,2])
|
||
b = np.array([0,10,0])
|
||
|
||
np.logical_and(a, b)
|
||
|
||
```
|
||
|
||
Out[8]:
|
||
|
||
```py
|
||
array([False, True, False], dtype=bool)
|
||
```
|
||
|
||
`0` 被认为是 `False`，非零则是 `True`。
|
||
|
||
比特操作：
|
||
|
||
In [9]:
|
||
|
||
```py
|
||
a = np.array([1,2,4,8])
|
||
b = np.array([16,32,64,128])
|
||
|
||
a | b
|
||
|
||
```
|
||
|
||
Out[9]:
|
||
|
||
```py
|
||
array([ 17, 34, 68, 136])
|
||
```
|
||
|
||
取反：
|
||
|
||
In [10]:
|
||
|
||
```py
|
||
a = np.array([1,2,3,4], np.uint8)
|
||
~a
|
||
|
||
```
|
||
|
||
Out[10]:
|
||
|
||
```py
|
||
array([254, 253, 252, 251], dtype=uint8)
|
||
```
|
||
|
||
左移：
|
||
|
||
In [11]:
|
||
|
||
```py
|
||
a << 3
|
||
|
||
```
|
||
|
||
Out[11]:
|
||
|
||
```py
|
||
array([ 8, 16, 24, 32], dtype=uint8)
|
||
```
|
||
|
||
要注意的是 `&` 的运算优先于比较运算如 `>` 等，所以必要时候需要加上括号：
|
||
|
||
In [12]:
|
||
|
||
```py
|
||
a = np.array([1,2,4,8])
|
||
b = np.array([16,32,64,128])
|
||
|
||
(a > 3) & (b < 100)
|
||
|
||
```
|
||
|
||
Out[12]:
|
||
|
||
```py
|
||
array([False, False, True, False], dtype=bool)
|
||
``` |