How to apply a conditional ffill fillna() with groupby dataframe

I have a dataframe with several classes/groups of data. One column indicates (True/False) whether the NaN contents of a specific row should have fillna() applied.

import pandas as pd
import numpy as np
df = pd.DataFrame(
            [["Bill", False, 13, 10, 15],
            ["Jane", False, 63, 17, 95],
            ["Bill", True, np.nan, 5, np.nan],
            ["Mary", False, 65, 13, np.nan]],
           columns=['Person','result','data1','data2','data3'])

My goal is to fill the NaN values for Bill with 13 and 15 respectively, because of the "True" parameter, while leaving other NaNs and data alone.
I've found solutions to simpler versions of this problem, but can't quite get it to work with the groupby.

This approach seems to lose the ability to reference the result column (KeyError 'result'):

    df2 = df.groupby('Person').transform(lambda x: x.fillna(method='ffill', axis=1) if x['result']==True else x) 

And this approach appears to give a similar problem:

    df2 = df.apply(
    lambda row: row.fillna(method='ffill') if row['result']==True else row
    )

Any advice?

The first approach almost works! Once you're in the groupby, you need to use apply, instead of transform, since you want the fillna to be applied over the all group. Also, you want the axis to be 0, not 1, otherwise it will fill from the adjacent columns

So, it should look something like:

df.groupby('Person').apply(lambda x: x.fillna(method='ffill', axis=0))

Notice that, this doesn't take into account your filter on "result". To take that into account, you can simply replace only the rows you want in the source database:

df[df["result"]] = df.groupby('Person', group_keys=False).apply(lambda x: x.fillna(method='ffill', axis=0))

Or, if you want the result to be in a second dataframe,

df2 = df.copy()
df2[df2["result"]] = df.groupby('Person', group_keys=False).apply(lambda x: x.fillna(method='ffill', axis=0))
df2