Question

我在Pandas日期框架中有小型市场产品销售数据：

date     product  name    turnover  qty   receipts_qty  profit  

2015-42  2889220  milk    17.73     3.000       3       13.2700  
2015-42  2905772  bread   40.94     2.000       2       40.9400 
2015-42  2876012  salt    291.36    5.000       3       292.2450  
2015-42  2846814  cakes   92.43     4.000       4       55.5300
...
2015-46  2889220  milk    12.44     2.000       2       9.9344
2015-46  2905772  bread   62.11     3.000       3       62.1100 
2015-42  2876012  salt    210.76    4.000       3       190.0050  
2015-42  2846814  cakes   120.27    5.000       4       72.1300


[14934 rows x 7 columns]

我需要的是在开始日期和结束日期之间根据'qty'列获得额外的两个数据帧 - 一个用于销售增长的产品，另一个用于销售额下降的产品。像这样：

Increase
name    qty_change  profit_change

bread   1.000       21.2700
cakes   1.000       16.6000
...

Decrease
name    qty_change  profit_change

milk    -1.000      -3,3356
salt    -1.000      -102,2400
...

Answer 1

您可以尝试：

import pandas as pd

products = pd.DataFrame({'product': [2889220, 2905772, 2876012, 2846814, 2889220, 2905772, 2876012, 2846814], 
    'date': ['2015-42', '2015-42', '2015-42', '2015-42', '2015-46', '2015-46', '2015-42', '2015-42'], 
    'receipts_qty': [3, 2, 3, 4, 2, 3, 3, 4], 
    'qty': [3.0, 2.0, 5.0, 4.0, 2.0, 3.0, 4.0, 5.0], 
    'profit': [13.27, 40.94, 292.245, 55.53, 9.9344, 62.11, 190.005, 72.13], 
    'name': ['milk', 'bread', 'solt', 'cackes', 'milk', 'bread', 'solt', 'cackes'], 
    'turnover': [17.73, 40.94, 291.36, 92.43, 12.44, 62.11, 210.76, 120.27]})


products[['qty_change','profit_change']] = products.groupby('name')[['qty','profit']].diff()

increase = products[products.qty_change > 0][['name','qty_change','profit_change']]
print(increase)

#     name  qty_change  profit_change
#5   bread         1.0          21.17
#7  cackes         1.0          16.60


decrease = products[products.qty_change < 0][['name','qty_change','profit_change']]

print(decrease)
#   name  qty_change  profit_change
#4  milk        -1.0        -3.3356
#6  solt        -1.0      -102.2400

我希望这会有所帮助。

在现有的基础上创建其他数据帧

1 个答案: