I want to create (for example) a small multi-index dataframe with 3 level index and one column of values. There will only be a handful of rows.

I want to create (for example) a small multi-index dataframe with 3 level index and one column of values. There will only be a handful of rows.

A requirement is that the data has to be easy to enter, a row at a time, from left to right (not top to bottom), so my first step is to make a list of lists.

Then I can use DataFrame method and then the set-index method like this:

data=[["Machine","Car","Sedan",10], ["Machine","Bike","2 Wheel",5], ["Machine","Bike","3 Wheel",4], ["Animal","Donkey","Big",2]]
column_names=["Category","Sub Category","Sub Sub Category","Value"]
df=pd.DataFrame(data,columns=column_names)
df.set_index(["Category","Sub Category","Sub Sub Category"])

But is there a way to make the multi-index DataFrame directly from a list, or something like it?

• python
• pandas
• dataframe
• multi-index

• Please remember that Stack Overflow is not your favourite Python forum, but rather a question and answer site for all programming related questions. Thus, always include the tag of the language you are programming in, that way other users familiar with that language can more easily find your question. Take the tour and read up on How to Ask to get more information on how this site works, then edit the question with the relevant tags. – Adriaan Commented Mar 14 at 8:38

1 Answer 1

You can create a MultiIndex from a list of tuples (or list of lists) with pandas.MultiIndex.from_tuples:

pd.MultiIndex.from_tuples([lst[:3] for lst in data],
                          names=['Category', 'Sub Category', 'Sub Sub Category']
                         )

Output:

MultiIndex([('Machine',    'Car',   'Sedan'),
            ('Machine',   'Bike', '2 Wheel'),
            ('Machine',   'Bike', '3 Wheel'),
            ( 'Animal', 'Donkey',     'Big')],
           names=['Category', 'Sub Category', 'Sub Sub Category'])

To create a DataFrame, you could split the sublists into index/values and pass them to the DataFrame constructor:

*index, values = map(list, zip(*data))
df = (pd.DataFrame({'Value': values}, index=index)
        .rename_axis(['Category', 'Sub Category', 'Sub Sub Category'])
     )

Output:

                                        Value
Category Sub Category Sub Sub Category       
Machine  Car          Sedan                10
         Bike         2 Wheel               5
                      3 Wheel               4
Animal   Donkey       Big                   2

Your approach is also perfectly valid, you could just reuse the variable with the list of columns to use as index and chain the operations:

df = (pd.DataFrame(data, columns=column_names)
        .set_index(column_names[:3])
     )