pandas中的数值计算及统计基础

 import pandas as pd
 import numpy as np

 df = pd.DataFrame({
     'key1': [4, 5, 3, np.nan, 2],
     'key2': [1, 2, np.nan, 4, 5],
     'key3': [1, 2, 3, 'j', 'k']
 }, index=['a', 'b', 'c', 'd', 'e'])
 print(df)
 print(df['key1'].dtype,df['key2'].dtype,df['key3'].dtype)
 print('-------')
 '''
    key1  key2 key3
 a   4.0   1.0    1
 b   5.0   2.0    2
 c   3.0   NaN    3
 d   NaN   4.0    j
 e   2.0   5.0    k
 float64 float64 object
 -------
 '''
 # 计算每一列的均值 df.mean()
 # 只统计数字列，默认忽略nan。
 print(df.mean())
 '''
 key1    3.5
 key2    3.0
 dtype: float64
 '''
 # 不忽略nan值计算均值
 # skipna默认为True，如果为False，有NaN的列统计结果仍为NaN
 m3 = df.mean(skipna=False)
 print(m3)
 '''
 key1   NaN
 key2   NaN
 dtype: float64
 '''
 # 计算单一列的均值
 print('计算单一列的均值',df['key2'].mean())
 '''
 计算单一列的均值 3.0
 '''

 df2 = pd.DataFrame({
     'key1': [1, 3, 5],
     'key2': [2, 4, 6],
     'key3': [3, 5, 7]
 }, index=['a', 'b', 'c'])
 # print(df2)
 # print('--------df2')
 # 计算df2每一行的均值并将其结果添加到新的列
 df2['mean'] = df2.mean(axis=1)
 print(df2)
 '''
    key1  key2  key3  mean
 a     1     2     3   2.0
 b     3     4     5   4.0
 c     5     6     7   6.0
 '''

 # 统计非NaN值的数量  count()
 print(df)
 print('-'*6)
 print(df.count())
 '''
    key1  key2 key3
 a   4.0   1.0    1
 b   5.0   2.0    2
 c   3.0   NaN    3
 d   NaN   4.0    j
 e   2.0   5.0    k
 ------
 key1    4
 key2    4
 key3    5
 dtype: int64
 '''

 # 统计
 print(df)
 print('-' * 6)
 print('df的最小值',df.min())
 print('df的最大值',df.max())
 print('df的key2列的最大值',df['key2'].max())
 print('统计df的分位数，参数q确定位置',df.quantile(q=0.75))
 print('对df求和',df.sum())
 print('求df的中位数，median(),50%分位数',df.median())
 print('求df的标准差，std()',df.std())
 print('求df的方差，var()',df.var())
 print('求skew样本的偏度,skew()',df.skew())
 print('求kurt样本的峰度，kurt()',df.kurt())
 print('df累计求和，cumsum()',df['key2'].cumsum())
 print('df累计求积，cumprod()',df['key2'].cumprod())
 print('求df的累计最大值，cummax()', df['key2'].cummax())
 print('求df的累计最小值，cummin()', df['key2'].cummin())
 '''
    key1  key2 key3
 a   4.0   1.0    1
 b   5.0   2.0    2
 c   3.0   NaN    3
 d   NaN   4.0    j
 e   2.0   5.0    k
 ------
 df的最小值 key1    2.0
 key2    1.0
 dtype: float64
 df的最大值 key1    5.0
 key2    5.0
 dtype: float64
 df的key2列的最大值 5.0
 统计df的分位数，参数q确定位置 key1    4.25
 key2    4.25
 Name: 0.75, dtype: float64
 对df求和 key1    14.0
 key2    12.0
 dtype: float64
 求df的中位数，median(),50%分位数 key1    3.5
 key2    3.0
 dtype: float64
 求df的标准差，std() key1    1.290994
 key2    1.825742
 dtype: float64
 求df的方差，var() key1    1.666667
 key2    3.333333
 dtype: float64
 求skew样本的偏度,skew() key1    0.0
 key2    0.0
 dtype: float64
 求kurt样本的峰度，kurt() key1   -1.2
 key2   -3.3
 dtype: float64
 df累计求和，cumsum() a     1.0
 b     3.0
 c     NaN
 d     7.0
 e    12.0
 Name: key2, dtype: float64
 df累计求积，cumprod() a     1.0
 b     2.0
 c     NaN
 d     8.0
 e    40.0
 Name: key2, dtype: float64
 求df的累计最大值，cummax() a    1.0
 b    2.0
 c    NaN
 d    4.0
 e    5.0
 Name: key2, dtype: float64
 求df的累计最小值，cummin() a    1.0
 b    1.0
 c    NaN
 d    1.0
 e    1.0
 Name: key2, dtype: float64
 '''

 # 唯一值 ：unique()
 s = pd.Series(list('kjdhsakjdhjfh'))
 sq = s.unique()
 print(s)
 print(sq)
 print('sq的类型：',type(sq))
 print('对sq进行重新排序：',pd.Series(sq).sort_values())
 '''
 0     k
 1     j
 2     d
 3     h
 4     s
 5     a
 6     k
 7     j
 8     d
 9     h
 10    j
 11    f
 12    h
 dtype: object
 ['k' 'j' 'd' 'h' 's' 'a' 'f']
 sq的类型： <class 'numpy.ndarray'>
 对sq进行重新排序： 5    a
 2    d
 6    f
 3    h
 1    j
 0    k
 4    s
 dtype: object
 '''
 # 对某一列进行值的计数，只能对一列，不能对Dataframe
 print(df['key2'].value_counts())

 # 判断Dataframe中的每个元素是否都是在某个列表中
 print(df)
 df_isin = df.isin([1,3])
 print(df_isin)
 '''
    key1  key2 key3
 a   4.0   1.0    1
 b   5.0   2.0    2
 c   3.0   NaN    3
 d   NaN   4.0    j
 e   2.0   5.0    k

     key1   key2   key3
 a  False   True   True
 b  False  False  False
 c   True  False   True
 d  False  False  False
 e  False  False  False
 '''