Pandas convert Dataframe to Nested Json

Question

My question is essentially the opposite of this one:

Create a Pandas DataFrame from deeply nested JSON

I'm wondering if it's possible to do the reverse. Given a table like:

     Library  Level           School Major  2013 Total
200  MS_AVERY  UGRAD  GENERAL STUDIES  GEST        5079
201  MS_AVERY  UGRAD  GENERAL STUDIES  HIST           5
202  MS_AVERY  UGRAD  GENERAL STUDIES  MELC           2
203  MS_AVERY  UGRAD  GENERAL STUDIES  PHIL          10
204  MS_AVERY  UGRAD  GENERAL STUDIES  PHYS           1
205  MS_AVERY  UGRAD  GENERAL STUDIES  POLS          53

Is it possible to generate a nested dict (or JSON) like:

dict:

{'MS_AVERY': 
    { 'UGRAD' :
        {'GENERAL STUDIES' : {'GEST' : 5}
                             {'MELC' : 2}

 ...

Answer 1

It seems not hard to create a function will build the recursive dictionary given your DataFrame object:

def fdrec(df):
    drec = dict()
    ncols = df.values.shape[1]
    for line in df.values:
        d = drec
        for j, col in enumerate(line[:-1]):
            if not col in d.keys():
                if j != ncols-2:
                    d[col] = {}
                    d = d[col]
                else:
                    d[col] = line[-1]
            else:
                if j!= ncols-2:
                    d = d[col]
    return drec

which will produce:

{'MS_AVERY':
    {'UGRAD':
        {'GENERAL STUDIES': {'PHYS': 1L, 
                             'POLS': 53L,
                             'PHIL': 10L,
                             'HIST': 5L,
                             'MELC': 2L,
                             'GEST': 5079L}}}}

Answer 2

Here's a solution I came up while working on this question:

def rollup_to_dict_core(x, values, columns, d_columns=None):
    if d_columns is None:
        d_columns = []

    if len(columns) == 1:
        if len(values) == 1:
            return x.set_index(columns)[values[0]].to_dict()
        else:
            return x.set_index(columns)[values].to_dict(orient='index')
    else:
        res = x.groupby([columns[0]] + d_columns).apply(lambda y: rollup_to_dict_core(y, values, columns[1:]))
        if len(d_columns) == 0:
            return res.to_dict()
        else:
            res.name = columns[1]
            res = res.reset_index(level=range(1, len(d_columns) + 1))
            return res.to_dict(orient='index')

def rollup_to_dict(x, values, d_columns=None):
    if d_columns is None:
        d_columns = []

    columns = [c for c in x.columns if c not in values and c not in d_columns]
    return rollup_to_dict_core(x, values, columns, d_columns)

>>> pprint(rollup_to_dict(df, ['2013 Total']))
{'MS_AVERY': {'UGRAD': {'GENERAL STUDIES': {'GEST': 5079,
                                            'HIST': 5,
                                            'MELC': 2,
                                            'PHIL': 10,
                                            'PHYS': 1,
                                            'POLS': 53}}}}