将行作为列索引、将列作为行索引的 Pythonic 方法

Question

我想改变数据框对象。我想将第一行作为列索引。第一列作为行索引。

import pandas as pd

wiki = "https://en.wikipedia.org/wiki/List_of_state_and_union_territory_capitals_in_India"
df = pd.read_html(wiki)[1]
df2 = df.copy()
df2.head()

目前我正在这样做(我在其中丢失了行索引名称):

df2.columns = df.iloc[0]
df2.drop(0, inplace=True)
df2.drop('No.', axis=1, inplace=True)
df2.head()

如何以更 Pythonic 的方式保留行索引名称？

Answer 1

您可以直接在 read_html 中指定您的意愿，使用 header 指定要使用哪一行作为列，以及 index_col 使用哪一列作为索引:

In [16]: df = pd.read_html(wiki,header=0,index_col=0)[1]

In [17]: df.head()
Out[17]: 
        State or union territory Administrative capitals Legislative capitals  \
No.                                                                             
1    Andaman and Nicobar Islands              Port Blair           Port Blair   
2                 Andhra Pradesh            Hyderabad[a]            Hyderabad   
3              Arunachal Pradesh                Itanagar             Itanagar   
4                          Assam                  Dispur             Guwahati   
5                          Bihar                   Patna                Patna   

    Judiciary capitals Year capital was established       The Former capital  
No.                                                                           
1              Kolkata                         1955     Calcutta (1945–1956)  
2            Hyderabad                         1959      Kurnool (1953-1956)  
3             Guwahati                         1986                      NaN  
4             Guwahati                         1975  Shillong[b] (1874–1972)  
5                Patna                         1912                      NaN