Python 如何基于字典通过共享列合并两个数据帧？

所以我有两个数据帧，我正试图通过一列连接。但是，这一公共列在两个数据帧之间是不同的，其中一个数据帧中的值是全名，而另一个数据帧中的值是缩写。我正在尝试制作某种映射字典，将名称等同于它们的缩写，这样数据帧就可以通过这个公共列直接连接起来。我将在下面进一步解释

我有这个数据帧（A）：

然后我得到了这个数据帧（B）：

现在我想创建以下数据帧（C）：

我会进一步解释。我有建筑物的数据，显示了3种类型的建筑物对应的3种类型的二氧化碳和氮氧化物排放量（数据框A）。然后，我有数据显示，在每个列出的邮政编码（数据框B）中，这些类型的房屋有多少。我最终想要创建一个数据框，显示每个邮政编码的二氧化碳和氮氧化物的总排放量（数据框C）。获取数据框C背后的想法是，我想生成一个数据框，然后通过邮政编码将其连接到GIS形状文件，这样我就可以映射每个邮政编码中的CO2和NOx排放量，并获得一个要连接的空白zipcodes形状文件。（我意识到绘制排放量图要比这复杂得多，但在我的项目的这一阶段，我保持这一简单）

所以我要做的是在“Building”列中将数据帧A连接到数据帧B。但问题是，数据框A中的“Building”列具有完整的名称，而数据框B中的“Building”列具有作为缩写的名称。我在想，我需要创建某种字典，将全名与要引用的缩写匹配，但我不确定如何将其放在这里

这可以用python实现吗？或者这真的比我想象的要复杂吗？我花了好几个小时试图思考如何合并这两个数据帧，但每次我都会更加困惑。我在概念化这段代码时遇到了很多困难，即使目标看起来很简单。我将非常感谢您的帮助和指导！抱歉让这些数据帧太长，但我觉得有必要捕获数据的结构/复杂性

---

谢谢大家!

这当然是可能的。您确实可以首先使用自定义函数将建筑物名称转换为缩写，该函数在连字符上拆分，并取每个单词的第一个字母。然后可以合并

建筑

和

样式

上的数据帧。最后，您可以在

Zip code

上查看

groupby

：

import pandas as pd

data1 = [ { "Building": "Multi-Family", "Style": "A", "CO2": 34, "NOx": 55 }, { "Building": "Multi-Family", "Style": "B", "CO2": 43, "NOx": 44 }, { "Building": "Multi-Family", "Style": "C", "CO2": 33, "NOx": 35 }, { "Building": "Single-Family", "Style": "A", "CO2": 34, "NOx": 26 }, { "Building": "Single-Family", "Style": "B", "CO2": 22, "NOx": 26 }, { "Building": "Single-Family", "Style": "C", "CO2": 65, "NOx": 48 }, { "Building": "Single-Family", "Style": "D", "CO2": 55, "NOx": 74 }, { "Building": "Studio", "Style": "A", "CO2": 46, "NOx": 35 }, { "Building": "Studio", "Style": "B", "CO2": 54, "NOx": 67 }, { "Building": "Studio", "Style": "C", "CO2": 57, "NOx": 58 } ]
data2 = [ { "Building": "MF", "Style": "A", "Zip_code": 11111, "Number": 4 }, { "Building": "MF", "Style": "A", "Zip_code": 22222, "Number": 3 }, { "Building": "MF", "Style": "A", "Zip_code": 33333, "Number": 2 }, { "Building": "MF", "Style": "B", "Zip_code": 11111, "Number": 1 }, { "Building": "MF", "Style": "B", "Zip_code": 22222, "Number": 1 }, { "Building": "MF", "Style": "C", "Zip_code": 22222, "Number": 1 }, { "Building": "MF", "Style": "C", "Zip_code": 33333, "Number": 6 }, { "Building": "SF", "Style": "A", "Zip_code": 11111, "Number": 7 }, { "Building": "SF", "Style": "A", "Zip_code": 22222, "Number": 5 }, { "Building": "SF", "Style": "B", "Zip_code": 44444, "Number": 3 }, { "Building": "SF", "Style": "B", "Zip_code": 55555, "Number": 8 }, { "Building": "SF", "Style": "B", "Zip_code": 66666, "Number": 6 }, { "Building": "SF", "Style": "C", "Zip_code": 11111, "Number": 9 }, { "Building": "SF", "Style": "C", "Zip_code": 22222, "Number": 9 }, { "Building": "ST", "Style": "A", "Zip_code": 33333, "Number": 3 }, { "Building": "ST", "Style": "A", "Zip_code": 44444, "Number": 5 }, { "Building": "ST", "Style": "B", "Zip_code": 55555, "Number": 5 }, { "Building": "ST", "Style": "B", "Zip_code": 66666, "Number": 3 }, { "Building": "ST", "Style": "C", "Zip_code": 11111, "Number": 2 }, { "Building": "ST", "Style": "C", "Zip_code": 22222, "Number": 9 }, { "Building": "ST", "Style": "C", "Zip_code": 33333, "Number": 1 } ]
df1 = pd.DataFrame(data1)
df2 = pd.DataFrame(data2)

def shorten_buildings(text):
    text = ''.join([i[0] for i in text.split('-')])
    return text if text != 'S' else 'ST'

df1['Building'] = df1['Building'].apply(shorten_buildings)

df2 = df2.merge(df1,  how='left', on=['Building','Style'])

df2['CO2'] = df2['Number'] * df2['CO2'] #total CO2
df2['NOx'] = df2['Number'] * df2['NOx'] #total NOx
df2.groupby(['Zip_code']).sum().drop('Number', axis=1)

输出：

邮政编码 二氧化碳 氮氧化物 11111 1116 994 22222 1446 1328 33333 461 483 44444 296 253 55555 446 543 66666 294 357

Building   Style    Zip_code     Number
---------------------------------------
      MF       A       11111         4 
      MF       A       22222         3
      MF       A       33333         2
      MF       B       11111         1
      MF       B       22222         1
      MF       C       22222         1 
      MF       C       33333         6
      SF       A       11111         7 
      SF       A       22222         5
      SF       B       44444         3
      SF       B       55555         8 
      SF       B       66666         6
      SF       C       11111         9
      SF       C       22222         9
      ST       A       33333         3
      ST       A       44444         5
      ST       B       55555         5
      ST       B       66666         3
      ST       C       11111         2 
      ST       C       22222         9 
      ST       C       33333         1

Zip_code      CO2    NOx
-------------------------
11111           ?      ?
22222           ?      ?
33333           ?      ?
44444           ?      ?
55555           ?      ?
66666           ?      ?

import pandas as pd

data1 = [ { "Building": "Multi-Family", "Style": "A", "CO2": 34, "NOx": 55 }, { "Building": "Multi-Family", "Style": "B", "CO2": 43, "NOx": 44 }, { "Building": "Multi-Family", "Style": "C", "CO2": 33, "NOx": 35 }, { "Building": "Single-Family", "Style": "A", "CO2": 34, "NOx": 26 }, { "Building": "Single-Family", "Style": "B", "CO2": 22, "NOx": 26 }, { "Building": "Single-Family", "Style": "C", "CO2": 65, "NOx": 48 }, { "Building": "Single-Family", "Style": "D", "CO2": 55, "NOx": 74 }, { "Building": "Studio", "Style": "A", "CO2": 46, "NOx": 35 }, { "Building": "Studio", "Style": "B", "CO2": 54, "NOx": 67 }, { "Building": "Studio", "Style": "C", "CO2": 57, "NOx": 58 } ]
data2 = [ { "Building": "MF", "Style": "A", "Zip_code": 11111, "Number": 4 }, { "Building": "MF", "Style": "A", "Zip_code": 22222, "Number": 3 }, { "Building": "MF", "Style": "A", "Zip_code": 33333, "Number": 2 }, { "Building": "MF", "Style": "B", "Zip_code": 11111, "Number": 1 }, { "Building": "MF", "Style": "B", "Zip_code": 22222, "Number": 1 }, { "Building": "MF", "Style": "C", "Zip_code": 22222, "Number": 1 }, { "Building": "MF", "Style": "C", "Zip_code": 33333, "Number": 6 }, { "Building": "SF", "Style": "A", "Zip_code": 11111, "Number": 7 }, { "Building": "SF", "Style": "A", "Zip_code": 22222, "Number": 5 }, { "Building": "SF", "Style": "B", "Zip_code": 44444, "Number": 3 }, { "Building": "SF", "Style": "B", "Zip_code": 55555, "Number": 8 }, { "Building": "SF", "Style": "B", "Zip_code": 66666, "Number": 6 }, { "Building": "SF", "Style": "C", "Zip_code": 11111, "Number": 9 }, { "Building": "SF", "Style": "C", "Zip_code": 22222, "Number": 9 }, { "Building": "ST", "Style": "A", "Zip_code": 33333, "Number": 3 }, { "Building": "ST", "Style": "A", "Zip_code": 44444, "Number": 5 }, { "Building": "ST", "Style": "B", "Zip_code": 55555, "Number": 5 }, { "Building": "ST", "Style": "B", "Zip_code": 66666, "Number": 3 }, { "Building": "ST", "Style": "C", "Zip_code": 11111, "Number": 2 }, { "Building": "ST", "Style": "C", "Zip_code": 22222, "Number": 9 }, { "Building": "ST", "Style": "C", "Zip_code": 33333, "Number": 1 } ]
df1 = pd.DataFrame(data1)
df2 = pd.DataFrame(data2)

def shorten_buildings(text):
    text = ''.join([i[0] for i in text.split('-')])
    return text if text != 'S' else 'ST'

df1['Building'] = df1['Building'].apply(shorten_buildings)

df2 = df2.merge(df1,  how='left', on=['Building','Style'])

df2['CO2'] = df2['Number'] * df2['CO2'] #total CO2
df2['NOx'] = df2['Number'] * df2['NOx'] #total NOx
df2.groupby(['Zip_code']).sum().drop('Number', axis=1)