Pandas：如何找到两列之间的差异

要查找 pandas DataFrame 中两列之间的差异，可以使用以下语法：

 df[' difference '] = df[' column1 '] - df[' column2 ']

以下示例展示了如何在实践中使用此语法。

示例 1：查找两列之间的差异

假设我们有以下 pandas DataFrame，显示两个区域（A 和 B）连续八个销售周期的总销售额：

 import pandas as pd

#createDataFrame
df = pd. DataFrame ({' period ': [1, 2, 3, 4, 5, 6, 7, 8],
                   ' A_sales ': [12, 14, 15, 13, 18, 20, 19, 24],
                   ' B_sales ': [14, 19, 20, 22, 24, 20, 17, 23]})

#view DataFrame
df

period A_sales B_sales
0 1 12 14
1 2 14 19
2 3 15 20
3 4 13 22
4 5 18 24
5 6 20 20
6 7 19 17
7 8 24 23

下面的代码展示了如何计算每个销售周期的B区和A区销售额之间的差异：

 #add new column to represent difference between B sales and A sales
df[' diff '] = df[' B_sales '] - df[' A_sales ']

#view DataFrame
df

        period A_sales B_sales diff
0 1 12 14 2
1 2 14 19 5
2 3 15 20 5
3 4 13 22 9
4 5 18 24 6
5 6 20 20 0
6 7 19 17 -2
7 8 24 23 -1

我们还可以使用pandas.Series.abs()函数计算销售额的绝对差异：

 #add new column to represent absolute difference between B sales and A sales
df[' diff '] = pd. Series . abs (df[' B_sales '] - df[' A_sales '])

#view DataFrame
df

	period A_sales B_sales diff
0 1 12 14 2
1 2 14 19 5
2 3 15 20 5
3 4 13 22 9
4 5 18 24 6
5 6 20 20 0
6 7 19 17 2
7 8 24 23 1

示例2：根据条件查找列之间的差异

我们还可以过滤 DataFrame 以仅显示列之间的差异小于或大于特定值的行。

例如，以下代码仅返回区域 A 的销售额大于区域 B 的销售额的行：

 #add new column to represent difference between B sales and A sales
df[' diff '] = df[' B_sales '] - df[' A_sales ']

#display rows where sales in region A is greater than sales in region B
df[df[' diff ']< 0 ]


        period A_sales B_sales diff
6 7 19 17 -2
7 8 24 23 -1

其他资源

Pandas：如何找到两条线之间的差异
Pandas：如何跨多列进行分组和聚合