Can I create a nested column pandas dataframe using a nested dictionary?

I am trying to think out of the box here, and my idea might be really bad. Feel free to point out better alternatives. I want to create a nested column pandas dataframe, both for visualization and analysis purposes. The output should look like this –

           Marks
  Physics   |      Biology
 Theo|Prac  |     Theo|Prac
  99 | 100  |      89 | 100 

My data is stored in the form of a nested dictionary –

nested_dict = {
    'Marks': {
        'Physics': {
            'Theo': 99,
            'Prac': 100
        },
        'Biology': {
            'Theo': 89,
            'Prac': 100
        }
    }
}

I think the above table looks great for visualization but I am not sure if it’ll make things easy for analysis. For analysis, I’d need to do operations on the subgroups, For eg: Physics percentage = (Theo + Prac)/200*100. Is a nested panda dataframe the best way to do the analysis?

Is there a way I can do that? Doing pd.DataFrame.from_dict(nested_dict) doesn’t seem to work. This is what I get –

                             Marks
Biology  {'Theo': 89, 'Prac': 100}
Physics  {'Theo': 99, 'Prac': 100}

>Solution :

Un-nesting can be achieved manually with concat and a dictionary comprehension, then unstack+to_frame and transposition:

out = (pd.concat({k: pd.DataFrame(v)
                  for k, v in nested_dict.items()},
                 axis=1)
         .unstack().to_frame().T
      )

You could also use json_normalize and convert the columns to MultiIndex:

out = pd.json_normalize(nested_dict)
out.columns = out.columns.str.split('.', expand=True)

Output:

    Marks                  
  Physics      Biology     
     Theo Prac    Theo Prac
0      99  100      89  100