Beginner data mining-data exploration (7): the main data exploration function in python extended statistical feature function

The cumulative calculation statistical characteristic function in the pandas library:

1、cumsum()

Give the sum of the first 1, 2, ..., n numbers in turn

2 od buy ()

Give the product of the first 1, 2, ..., n numbers in turn

3、cummax()

Give the maximum value of the first 1, 2, ..., n numbers in turn

4、cummin()

Give the minimum of the first 1, 2, ..., n numbers in turn

note:

The cum series function appears as a method of the DataFrame or series object

Naming format: D.cumsm()

The rolling calculation statistical feature function in the pandas library:

1、rolling_sum()

Calculate the sum of data samples (calculated by column)

2、rolling_mean()

Calculate the arithmetic mean of the data sample

3 、 rolling_var ()

Calculate the variance of the data sample

4、rolling_std()

Calculate the standard deviation of the data sample

5、rolling_corr()

Calculate the spearman (Pearson) correlation coefficient matrix of the data sample

6 、 rolling_cov ()

Calculate the covariance matrix of the data sample

7、rolling_skew()

Calculate the skewness of the sample value (third moment)

8、rolling_kurt()

Calculate the kurtosis of the sample value (fourth moment)

note:

The rolling_ series function is a pandas function, not a method of the DataFrame or series object. Therefore, the format of the rolling_ series function is: pd.rolling_mean(Dk) , which means that the mean value is calculated once every k columns.

Examples:

import pandas as pd
D=pd.Series(range(0, 20))  # 构造Series，内容为0~19共20个整数
print(D)

print('前n项和:\n',D.cumsum())  # 给出前n项和

print('相邻两项求和:\n',D.rolling(2).sum())  # 依次对相邻两项求和