Numpy常用内置函数及运算函数

人工智障调包侠

于 2024-08-28 06:06:31 发布

阅读量279

点赞数 11

分类专栏： Numpy 文章标签： numpy python

本文链接：https://blog.csdn.net/bbaaa123/article/details/141617238

版权

Numpy 专栏收录该内容

3 篇文章 0 订阅

订阅专栏

常用运算函数

矩阵加减运算

矩阵加减元算分为相同形状矩阵间的加减以及广播机制运算

import numpy as np
from sympy.benchmarks.bench_discrete_log import data_set_3

data1 = np.array([[1,2,3],[4,5,6]])
data2 = np.array([[1,2,3]])
data3 = np.array([1,2,3])
print(data1+data2)
print(data1+data3)
print(data1-data2)
print(data1-data3)
# [[2 4 6]
#  [5 7 9]]
# [[2 4 6]
#  [5 7 9]]
# [[0 0 0]
#  [3 3 3]]
# [[0 0 0]
#  [3 3 3]]

上面代码演示的是通过实现广播机制来进行加减法运算,具体实现方式为,若两个数组维度不同,则会给维度较小的数据进行维度扩充,实现方式是将其最外层添加一层值为1的维度,直到与另一维度相同,再对值为1的维度进行复制,以此实现广播机制,若仍无法进行形状匹配,则无法实现广播机制

广播机制逐元素乘除

乘法可用*或multiply(),除法可用/或divide()

import numpy as np

data1 = np.array([[1,2,3],[4,5,6]])
print(data1*2)
print(np.multiply(data1,2))
print(data1/2)
print(np.divide(data1,2))
# [[ 2  4  6]
#  [ 8 10 12]]
# [[ 2  4  6]
#  [ 8 10 12]]
# [[0.5 1.  1.5]
#  [2.  2.5 3. ]]
# [[0.5 1.  1.5]
#  [2.  2.5 3. ]]

向量乘法运算

向量乘法运算指的是点乘计算,也就是[a,b]*[c,d]=a*b+c*d

import numpy as np

data1 = np.array([1,2,3])
data2 = np.array([4,5,6])
print(data1.dot(data2))
# 32

矩阵乘法运算

矩阵乘法运算分为对应位置元素相乘和矩阵乘法

对应位置元素相乘

对应位置元素相乘可用*或multiply实现

import numpy as np

data1 = np.array([[1,2,3],[4,5,6]])
data2 = np.array([[1,2,3],[4,5,6]])
print(np.multiply(data1,data2))
print(data1*data2)
# [[ 1  4  9]
#  [16 25 36]]
# [[ 1  4  9]
#  [16 25 36]]

矩阵乘法

矩阵乘法通过dot()函数实现,要求左乘的矩阵的列数等于被乘的矩阵的行数

import numpy as np

data1 = np.array([[1,2,3],[4,5,6]])
data2 = np.array([[1,2],[3,4],[5,6]])
print(data1.dot(data2))
# [[22 28]
#  [49 64]]

常用内置函数

基本函数

np.ceil()

np.ceil()用于数组元素向上取整

import numpy as np

data1 = np.random.randn(4,5)
print(data1)
print(np.ceil(data1))
# [[ 3.05585723 -0.37580912 -1.63515777  0.44520431 -1.25620038]
#  [-1.51531436  1.00796354 -2.95765468  0.33724942  0.79253138]
#  [ 0.2386308  -1.0385805  -0.82818326  1.61327862  0.86117067]
#  [-1.40692597 -1.65335482  0.25097572 -0.51377466  1.06806423]]
# [[ 4. -0. -1.  1. -1.]
#  [-1.  2. -2.  1.  1.]
#  [ 1. -1. -0.  2.  1.]
#  [-1. -1.  1. -0.  2.]]

np.floor()

np.floor()用于数组元素向下取整

import numpy as np

data1 = np.random.randn(4,5)
print(data1)
print(np.floor(data1))
# [[ 0.92378641  1.35349128  0.57975044  1.20847343  0.67943746]
#  [-1.19182311  0.57659921  0.60981519  1.42663667  0.49602901]
#  [-0.3224854   0.72609288  0.65194753  2.05265749 -1.11113541]
#  [-1.93028677 -1.34332758 -0.42339502 -0.46828546 -1.11569573]]
# [[ 0.  1.  0.  1.  0.]
#  [-2.  0.  0.  1.  0.]
#  [-1.  0.  0.  2. -2.]
#  [-2. -2. -1. -1. -2.]]

np.rint()

np.rint()用于数组元素四舍五入取整

import numpy as np

data1 = np.random.randn(4,5)
print(data1)
print(np.rint(data1))
# [[ 0.32973071 -0.12852531 -0.29609136  0.1099612  -1.32187404]
#  [-1.7301812   1.33769697 -0.08907079 -0.85909786  0.1095878 ]
#  [ 0.69617615  0.21219543 -1.54136    -1.21383603  1.22538225]
#  [-0.51724409 -0.06505319  0.23575689  0.12717578 -1.86922677]]
# [[ 0. -0. -0.  0. -1.]
#  [-2.  1. -0. -1.  0.]
#  [ 1.  0. -2. -1.  1.]
#  [-1. -0.  0.  0. -2.]]

np.isnan()

np.isnan()用于数组元素进行空值判断

import numpy as np

data1 = np.array([1,2,np.nan,3,np.nan,5])
print(np.isnan(data1))
# [False False  True False  True False]

np.where()

np.where()用于添加condition(条件)来进行元素筛选,返回对应元素下标,也可设置a(满足条件的返回值),b(不满足条件的返回值)

import numpy as np

data1 = np.arange(15)
print(np.where(data1 > 5))
print(np.where(data1 > 5,'True','False'))
# (array([ 6,  7,  8,  9, 10, 11, 12, 13, 14]),)
# ['False' 'False' 'False' 'False' 'False' 'False' 'True' 'True' 'True'
#  'True' 'True' 'True' 'True' 'True' 'True']

当where判别的数组为二维时,where会分别返回符合要求的元素的行下标和列下标,一一组合就是符合要求的元素的下标

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(np.where(data1>5))
# (array([1, 1, 1, 1, 2, 2, 2, 2, 2]), array([1, 2, 3, 4, 0, 1, 2, 3, 4]))

np.any()

np.any()的参数设置为判别条件,若所有元素中有任意元素满足,则范围True,否则返回False

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(np.any(data1>5))
# True

np.all()

np.all()的参数设置为判别条件,若所有元素都满足,则范围True,否则返回False

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(np.all(data1>5))
# False

统计函数

下述统计函数的演示都会用二维矩阵进行演示,每个函数都有一个参数axis,取值范围为[0,ndim)表示不同维度,在二维矩阵中,若axis取值为0,则是延行统计每一列,若取值为1,则是延列统计每一行

np.sum()

求和运算

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.sum())
print(data1.sum(axis=0))
print(data1.sum(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 105
# [15 18 21 24 27]
# [10 35 60]

np.min()

统计最小值

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.min())
print(data1.min(axis=0))
print(data1.min(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 0
# [0 1 2 3 4]
# [ 0  5 10]

np.max()

统计最大值

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.max())
print(data1.max(axis=0))
print(data1.max(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 14
# [10 11 12 13 14]
# [ 4  9 14]

np.mean()

统计平均值

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.mean())
print(data1.mean(axis=0))
print(data1.mean(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 7.0
# [5. 6. 7. 8. 9.]
# [ 2.  7. 12.]

np.argmax()

统计最大值下标

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.argmax())
print(data1.argmax(axis=0))
print(data1.argmax(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 14
# [2 2 2 2 2]
# [4 4 4]

np.argmin()

统计最小值下标

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.argmin())
print(data1.argmin(axis=0))
print(data1.argmin(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 0
# [0 0 0 0 0]
# [0 0 0]

np.cumsum()

按位置统计到当前位置为止的累加值

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.cumsum())
print(data1.cumsum(axis=0))
print(data1.cumsum(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# [  0   1   3   6  10  15  21  28  36  45  55  66  78  91 105]
# [[ 0  1  2  3  4]
#  [ 5  7  9 11 13]
#  [15 18 21 24 27]]
# [[ 0  1  3  6 10]
#  [ 5 11 18 26 35]
#  [10 21 33 46 60]]

np.cumprod()

按位置统计到当前位置为止的累乘值

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.cumprod())
print(data1.cumprod(axis=0))
print(data1.cumprod(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
# [[  0   1   2   3   4]
#  [  0   6  14  24  36]
#  [  0  66 168 312 504]]
# [[     0      0      0      0      0]
#  [     5     30    210   1680  15120]
#  [    10    110   1320  17160 240240]]

np.std()

统计标准差

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.std())
print(data1.std(axis=0))
print(data1.std(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 4.320493798938574
# [4.0824829 4.0824829 4.0824829 4.0824829 4.0824829]
# [1.41421356 1.41421356 1.41421356]

np.var()

统计方差

import numpy as np

data1 = np.arange(15).reshape(3, 5)
print(data1)
print(data1.var())
print(data1.var(axis=0))
print(data1.var(axis=1))

# [[ 0  1  2  3  4]
#  [ 5  6  7  8  9]
#  [10 11 12 13 14]]
# 18.666666666666668
# [16.66666667 16.66666667 16.66666667 16.66666667 16.66666667]
# [2. 2. 2.]