Pandas Dataframe Growing In Rows And Columns

I have the following dataframe (which may grow in rows and Info columns): City Country Info1 Info2 BCN Spain 3 5.6 Moscow Russia 4 7 I'm trying t

Solution 1:

If you are okay with a small tweak in your output, you can use melt and to_dict directly to get separate dictionaries for each info:

>>> df.melt(['City', 'Country']).to_dict('r')

[{'City': 'BCN', 'Country': 'Spain', 'value': 3.0, 'variable': 'Info1'},
 {'City': 'Moscow', 'Country': 'Russia', 'value': 4.0, 'variable': 'Info1'},
 {'City': 'BCN', 'Country': 'Spain', 'value': 5.6, 'variable': 'Info2'},
 {'City': 'Moscow', 'Country': 'Russia', 'value': 7.0, 'variable': 'Info2'}]

Solution 2:

For a non-Pandas-specific solution, this split_rows function works for any iterable of namedtuples (or if you change the rd = ... line, anything that can be dictified).

import pandas as pd


def split_rows(namedtuple_iterable, cols):
    for row in namedtuple_iterable:
        rd = row._asdict()
        cs = [(col, rd.pop(col)) for col in cols]
        for key, value in cs:
            yield {**rd, key: value}


df = pd.DataFrame(
    {
        "city": ["BCN", "Moscow"],
        "country": ["Spain", "Russia"],
        "inf_1": [3, 5],
        "inf_2": [4, 7],
    }
)


for sr in split_rows(df.itertuples(), ("inf_1", "inf_2")):
    print(sr)

outputs

{'Index': 0, 'city': 'BCN', 'country': 'Spain', 'inf_1': 3}
{'Index': 0, 'city': 'BCN', 'country': 'Spain', 'inf_2': 4}
{'Index': 1, 'city': 'Moscow', 'country': 'Russia', 'inf_1': 5}
{'Index': 1, 'city': 'Moscow', 'country': 'Russia', 'inf_2': 7}