How to Flatten a Hierarchical Column Index in Pandas?

Flattening Hierarchical Index Columns

To flatten a hierarchical index in the columns of a Pandas DataFrame, follow these steps:

1. Set Columns to Top Level:

df.columns = df.columns.get_level_values(0)

This will set the top level of the hierarchical index as the new column names.

2. Join MultiIndex into Single Index (Optional):

If you want to further combine the MultiIndex into a single index, you can do the following:

df.columns = [' '.join(col).strip() for col in df.columns.values]

This will join the column names using space as a separator, stripping any leading or trailing whitespace.

Example:

Consider the following DataFrame with a hierarchical index in columns:

df = pd.DataFrame({'s_PC': [1, 0, 1], 's_CL': [0, 0, 10]},
                    index = pd.MultiIndex.from_tuples([
                        ('day', 1),('day', 2),('day', 3)
                    ]), columns = pd.MultiIndex.from_tuples([
                        ('USAF', ''),('WBAN', ''),('year', 1993)
                    ]))

Applying the flattening operation:

# Set columns to top level
df.columns = df.columns.get_level_values(0)

# Join MultiIndex into single index
df.columns = [' '.join(col).strip() for col in df.columns.values]

Output:

   USAF  WBAN  year  day  s_PC  s_CL
0   702   265  1993   1     1     0
1   702   265  1993   2     0     0
2   702   265  1993   3     1    10

The hierarchical index has been flattened into a single index.

The above is the detailed content of How to Flatten a Hierarchical Column Index in Pandas?. For more information, please follow other related articles on the PHP Chinese website!