ENH: Adding an empty level at the end of a multi index

[HichemAK]

Feature Type

• Adding new functionality to pandas

• Changing existing functionality in pandas

• Removing existing functionality in pandas

Problem Description

Hello,

I apologize if this was discussed before, if the solution to this already exists, or if this is irrelevant.

Basically, I wish there was a simple function to add a level in a multi index (in my case, it is a multi index columns) at a given position. As of now the only solution I found was to recreate the multi index completely with an additional level (see below)

I find this useful to concatenate two dataframes (axis=1) with a different number of levels in the columns.

Feature Description

Add a new function to MultiIndex.insert_level(position : int, value : Any | Iterable[Any])

I just thought maybe supporting multiple values can be useful to specify exactly what the new level should contain. If there is only one value, fill all the level with this value.

I'm not sure about the arguments "position" and "value" although insert_level feels nice.

Alternative Solutions

new_columns = pd.MultiIndex.from_tuples(
    [col + ('',) if isinstance(col, tuple) else (col, '') 
     for col in df.columns]
)
df.columns = new_columns

Additional Context

No response

[Chiwendaiyue]

Hi @HichemAK I'm new here and I find the idea of adding an insert_level method to MultiIndex very useful for handling hierarchical indices.To make sure I understand your requirement correctly, could you please confirm if this is what you had in mind?

• Insert a new level at position 0 with a scalar value

• new_index = multi_index.insert_level(0, 'new_value')

• Insert a new level at position 1 with different values for each item

• new_index = multi_index.insert_level(1, ['X', 'Y', 'Z'])

• Insert with a specific name for the new level

• new_index = multi_index.insert_level(1, 'value', name='level_name')

I've implemented a prototype that addresses this functionality.