2021-03-29 python数据处理系统学习（十一）统计相关函数

2021-03-29 python数据处理系统学习（十一）统计基础的相关函数

axis=0表示对行作用得到列的性质，axis=1表示对列作用得到行的性质。
先读入数据，

import numpy as np
data =np.genfromtxt(r'C:\Users\wangyu\Desktop\毕业论文\python练习1.txt',delimiter='	')

data.shape
Out[6]: (13, 4)

读取结果如下：

1、求和

data.sum(axis=0)#对行进行作用，返回4列总和
Out[7]: array([1.120000e+03, 1.241450e+01, 2.709128e+07, 1.906630e+05])

data.sum(axis=1)#对列进行作用，返回13行总和
Out[8]: 
array([-82959690.6960142 ,  11620329.51070258,  13974025.23977928,
        42850879.00860883,   -682932.54615708,   1269070.3025655 ,
        12168307.98312435,   5124024.47889916,   7935968.65946828,
         9321502.0170493 ,    135926.29829416,   3564250.70368717,
         2961414.42229263])

沿着指定轴的元素累加和所组成的数组（cumsum）：

data.cumsum(axis=0)
Out[11]: 
array([[ 1.00000000e+02,  9.63985802e-01, -8.29709267e+07,
         1.11350000e+04],
       [ 2.00000000e+02,  1.92468839e+00, -7.13937981e+07,
         5.42350000e+04],
       [ 2.60000000e+02,  2.90446767e+00, -5.74478889e+07,
         8.22900000e+04],
       [ 3.20000000e+02,  3.82307649e+00, -1.45996328e+07,
         8.48520000e+04],
       [ 4.00000000e+02,  4.77971941e+00, -1.52857353e+07,
         8.79410000e+04],
       [ 5.00000000e+02,  5.72628491e+00, -1.40284929e+07,
         9.96680000e+04],
       [ 6.00000000e+02,  6.69940927e+00, -1.88096090e+06,
         1.20343000e+05],
       [ 7.00000000e+02,  7.64230843e+00,  3.21282664e+06,
         1.50479000e+05],
       [ 8.00000000e+02,  8.61877671e+00,  1.11387353e+07,
         1.60438000e+05],
       [ 8.80000000e+02,  9.55682601e+00,  2.04557804e+07,
         1.64814000e+05],
       [ 9.40000000e+02,  1.05095202e+01,  2.05878797e+07,
         1.68580000e+05],
       [ 1.02000000e+03,  1.14802073e+01,  2.41500295e+07,
         1.70600000e+05],
       [ 1.12000000e+03,  1.24145000e+01,  2.70912800e+07,
         1.90663000e+05]])

data.cumsum(axis=1)
Out[12]: 
array([[ 1.00000000e+02,  1.00963986e+02, -8.29708257e+07,
        -8.29596907e+07],
       [ 1.00000000e+02,  1.00960703e+02,  1.15772295e+07,
         1.16203295e+07],
       [ 6.00000000e+01,  6.09797793e+01,  1.39459702e+07,
         1.39740252e+07],
       [ 6.00000000e+01,  6.09186088e+01,  4.28483170e+07,
         4.28508790e+07],
       [ 8.00000000e+01,  8.09566429e+01, -6.86021546e+05,
        -6.82932546e+05],
       [ 1.00000000e+02,  1.00946566e+02,  1.25734330e+06,
         1.26907030e+06],
       [ 1.00000000e+02,  1.00973124e+02,  1.21476330e+07,
         1.21683080e+07],
       [ 1.00000000e+02,  1.00942899e+02,  5.09388848e+06,
         5.12402448e+06],
       [ 1.00000000e+02,  1.00976468e+02,  7.92600966e+06,
         7.93596866e+06],
       [ 8.00000000e+01,  8.09380493e+01,  9.31712602e+06,
         9.32150202e+06],
       [ 6.00000000e+01,  6.09526942e+01,  1.32160298e+05,
         1.35926298e+05],
       [ 8.00000000e+01,  8.09706872e+01,  3.56223070e+06,
         3.56425070e+06],
       [ 1.00000000e+02,  1.00934293e+02,  2.94135142e+06,
         2.96141442e+06]])

2、平均数

data.mean(axis=0)#对行进行作用，返回4列平均数
Out[9]: array([8.61538462e+01, 9.54961536e-01, 2.08394461e+06, 1.46663846e+04])

data.mean(axis=1)#对列进行作用，返回13行平均数
Out[10]: 
array([-20739922.67400355,   2905082.37767565,   3493506.30994482,
        10712719.75215221,   -170733.13653927,    317267.57564138,
         3042076.99578109,   1281006.11972479,   1983992.16486707,
         2330375.50426233,     33981.57457354,    891062.67592179,
          740353.60557316])

3、累积求积

data.cumprod()
Out[13]: 
array([ 1.00000000e+002,  9.63985802e+001, -7.99827953e+009,
       -8.90608425e+013, -8.90608425e+015, -8.55609815e+015,
       -9.90550481e+022, -4.26927257e+027, -2.56156354e+029,
       -2.50976689e+029, -3.50009814e+036, -9.81952532e+040,
       -5.89171519e+042, -5.41218154e+042, -2.31902541e+050,
       -5.94134309e+053, -4.75307447e+055, -4.54699506e+055,
        3.11970469e+061,  9.63676779e+064,  9.63676779e+066,
        9.12183194e+066,  1.14683535e+073,  1.34489381e+077,
        1.34489381e+079,  1.30874892e+079,  1.58980694e+086,
        3.28692585e+090,  3.28692585e+092,  3.09923962e+092,
        1.57868681e+099,  4.75753058e+103,  4.75753058e+105,
        4.64557771e+105,  3.68204247e+112,  3.66694610e+116,
        2.93355688e+118,  2.75182098e+118,  2.56388402e+125,
        1.12195565e+129,  6.73173387e+130,  6.41328352e+130,
        8.47190556e+135,  3.19051963e+139,  2.55241571e+141,
        2.47759718e+141,  8.82557213e+147,  1.78276557e+151,
        1.78276557e+153,  1.66562474e+153,  4.89901958e+159,
        9.82890298e+163])

4、最大值

data.max()
Out[14]: 42848256.09

data.max(axis=0)
Out[15]: array([1.00000000e+02, 9.79779283e-01, 4.28482561e+07, 4.31000000e+04])

data.max(axis=1)
Out[16]: 
array([1.11350000e+04, 1.15771286e+07, 1.39459093e+07, 4.28482561e+07,
       3.08900000e+03, 1.25724236e+06, 1.21475320e+07, 5.09378754e+06,
       7.92590868e+06, 9.31704508e+06, 1.32099346e+05, 3.56214973e+06,
       2.94125049e+06])

5、计算分位数

np.percentile(data,50)
Out[17]: 100.0

np.median(data)
Out[18]: 100.0

可以看出这两个公式表示的意思一样

np.percentile(data,[10,20,30,40,50])
Out[19]: 
array([  0.94326579,   0.96532608,  60.        ,  88.        ,
       100.        ])

6、极差

np.ptp(data)
Out[22]: 125819182.75

data.max()-data.min()
Out[23]: 125819182.75

np.ptp(data,axis=0)
Out[24]: array([4.00000000e+01, 6.11704630e-02, 1.25819183e+08, 4.10800000e+04])

7、判断大小

data>100
Out[25]: 
array([[False, False, False,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False, False,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True],
       [False, False,  True,  True]])

np.sum(data>100)
Out[26]: 24