1. import pandas as pd
    4. def csv_to_df(file):
    5.     stock_df = pd.read_csv(file)
    6.     # print(stock_df.head(10))
    7.     return stock_df
    10. def group_stock(stock_df):
    11.     stock_group = stock_df.groupby('Stock_code')
    12.     print(stock_group.describe())
    15. if __name__ == '__main__':
    16.     filename = '../data/stock_all.csv'
    17.     stu_df = csv_to_df(filename)
    18.     group_stock(stu_df)

    百分位数:统计学术语，如果将一组数据从小到大排序，并计算相应的累计百分位，则某一百分位所对应数据的值就称为这一百分位的百分位数。可表示为：一组n个观测值按数值大小排列。如，处于p%位置的值称第p百分位数

    1.               High                       ...  Adj_Close                      
    2.              count       mean       std  ...        50%        75%        max
    3. Stock_code                               ...                                 
    4. 600000      2633.0   9.558350  2.623207  ...   9.101338  10.760014  12.523858
    5. 600004      2632.0   9.440567  4.473545  ...   7.628153  12.611245  23.260876
    6. 600006      2632.0   5.204897  2.021496  ...   4.530794   5.727762  14.432583
    7. 600007      2633.0  13.849244  3.575867  ...  12.769957  15.113232  21.810545
    8. 600008      2630.0   3.976465  1.335394  ...   3.221414   3.978866   8.907692
    9. ...            ...        ...       ...  ...        ...        ...        ...
    10. 600395      2632.0  10.996351  5.393427  ...   6.807619  10.709707  17.755426
    11. 600396      2626.0   4.119981  1.786175  ...   3.616284   4.865564  13.316794
    12. 600397      2600.0   5.243262  2.258282  ...   4.890000   5.730211  14.460000
    13. 600398      2039.0   9.287401  3.447771  ...   8.984260  10.543828  19.097530
    14. 600400      2632.0   3.246414  1.737070  ...   3.225956   4.291482   7.573524
    16. [315 rows x 48 columns]

    统计各股票数据条数

    1. def group_stock(stock_df):
    2.     stock_group = stock_df.groupby('Stock_code')
    3.     print(stock_group.Stock_code.count())
    1. Stock_code
    2. 600000    2633
    3. 600004    2632
    4. 600006    2632
    5. 600007    2633
    6. 600008    2630
    7.           ... 
    8. 600395    2632
    9. 600396    2626
    10. 600397    2600
    11. 600398    2039
    12. 600400    2632
    13. Name: Stock_code, Length: 315, dtype: int64

    统计各股票收盘价的最高价

    1. def group_stock(stock_df):
    2.     stock_group = stock_df.groupby('Stock_code')
    3.     print(stock_group.Close.max())
    1. Stock_code
    2. 600000    14.055900
    3. 600004    23.500000
    4. 600006    15.190000
    5. 600007    23.750000
    6. 600008     9.785000
    7.             ...    
    8. 600395    26.200001
    9. 600396    13.630000
    10. 600397    14.460000
    11. 600398    19.959999
    12. 600400     8.246750
    13. Name: Close, Length: 315, dtype: float64

    每支股票每天最高价的历史最高价

    1. def group_stock(stock_df):
    2.     stock_group = stock_df.groupby('Stock_code')
    3.     print(stock_group.High.max())
    1. Stock_code
    2. 600000    14.069900
    3. 600004    23.690001
    4. 600006    16.350000
    5. 600007    23.959999
    6. 600008     9.840000
    7.             ...    
    8. 600395    26.733299
    9. 600396    14.370000
    10. 600397    15.000000
    11. 600398    21.059999
    12. 600400     8.246750
    13. Name: High, Length: 315, dtype: float64

    每支股票每天最高价的历史最高价和最低价

    1. def group_stock(stock_df):
    2.     stock_group = stock_df.groupby('Stock_code')
    3.     print(stock_group.High.agg([max, min]))
    1.                   max      min
    2. Stock_code                    
    3. 600000      14.069900  5.02098
    4. 600004      23.690001  4.15862
    5. 600006      16.350000  2.65000
    6. 600007      23.959999  8.81000
    7. 600008       9.840000  1.99500
    8. ...               ...      ...
    9. 600395      26.733299  4.89000
    10. 600396      14.370000  1.71000
    11. 600397      15.000000  1.73000
    12. 600398      21.059999  3.00000
    13. 600400       8.246750  1.05519
    15. [315 rows x 2 columns]