index_col can be used to specify one column to provide row indices. A list of columns can be passed as indices, which leads to multi-indexing. Let's look at an example:
pd.read_csv('mindex.txt')pd.read_csv('mindex.txt',index_col=[0,1])
Take a look at the following screenshot:
Single index (left) and multi-index (right) on the same data
This kind of multi-indexing makes it easy to subset based on either an index or both:
data.loc[1977]data.loc[(1977,'A')]
We get the following output:
Subsetting multi-indexed data ...
