时间戳

首页 > 解决方案 > 使用熊猫重塑/堆叠2列？

python - 使用熊猫重塑/堆叠2列？

问题描述

我对 Python/Pandas 比较陌生。我有一个按周列出的销售和单位产品列表，如下所示：

Product(index)    Desc1    Desc2    Desc3    Week1    Week2    Week3    Week1.1    Week2.1    Week3.1
Product1          Words1   Words1   Words1   $$$      $$$      $$$      ###        ###        ###
Product2          Words2   Words2   Words2   $$$      $$$      $$$      ###        ###        ###
Product3          Words3   Words3   Words3   $$$      $$$      $$$      ###        ###        ###

期望的结果是：

Product(index)    Desc1    Desc2    Desc3    Sales (By Week)    Units (By Week)
Product1          Words1   Words1   Words1   $$$                ###
Product1          Words1   Words1   Words1   $$$                ###
Product1          Words1   Words1   Words1   $$$                ###
.
.
.
ProductN          WordsN   WordsN   WordsN   $$$                ###
ProductN          WordsN   WordsN   WordsN   $$$                ###
ProductN          WordsN   WordsN   WordsN   $$$                ###

我尝试用多索引做一些事情：

iterables = [[Desc1, Desc2, Desc3],
            [Week1, Week2, Week3, Week1.1, Week2.1, Week3.1]]

test1 = pd.DataFrame(iterables, names=['Sales', 'Units'])
pd.MultiIndex.from_frame(test1)

附带说明一下，销售日期和单位日期是相同的（例如 8 月 1 日至 5 日），但是当 I 时，由于重复read_csv，它加载了第二列。.1这是一个问题吗？

标签： pythonpandas

解决方案

让我们尝试使用pandas.melt+pandas.concat

import pandas as pd

sales = (
    pd.melt(df, id_vars=['Product(index)', 'Desc1', 'Desc2', 'Desc3'],
              value_vars=['Week1', 'Week2', 'Week3'],
              value_name='Sales (By Week)')
)

unit = (
    pd.melt(df, id_vars=['Product(index)', 'Desc1', 'Desc2', 'Desc3'],
              value_vars=['Week1.1', 'Week2.1', 'Week3.1'],
              value_name='Units (By Week)')
)

print(pd.concat([sales, unit[['Units (By Week)']]], axis=1))

  Product(index)   Desc1   Desc2  ... variable Sales (By Week) Units (By Week)
0       Product1  Words1  Words1  ...    Week1             $$$             ###
1       Product2  Words2  Words2  ...    Week1             $$$             ###
2       Product3  Words3  Words3  ...    Week1             $$$             ###
3       Product1  Words1  Words1  ...    Week2             $$$             ###
4       Product2  Words2  Words2  ...    Week2             $$$             ###
5       Product3  Words3  Words3  ...    Week2             $$$             ###
6       Product1  Words1  Words1  ...    Week3             $$$             ###
7       Product2  Words2  Words2  ...    Week3             $$$             ###
8       Product3  Words3  Words3  ...    Week3             $$$             ###

推荐阅读