Match the values of two dataframes and return a dictionary-CodePudding

I have a dataframe like this

df1 = pd.DataFrame(
       {   
            "index":    ["December",  "December",  "November", "November"],
            "index1":   ["285",       "296",       "397",      "300"]})


     index     index1
0   December    285
1   December    296
2   November    397
3   November    300

And also o second dataframe like this

df2 = pd.DataFrame(
    {   
        "index":     ["December",  "December",  "November", "November"],
        "index1":    ["285",       "296",       "397",      "300"],
        "eventid":   ["67511",     "67512",     "64515",    "64517"]})


     index     index1    eventid
0   December    285       67511
1   December    296       67512
2   November    397       64515
3   November    300       64517

I want to match index & index1 of df1 and index & index1 of df2 and return the eventid of df2 in a dictionary

For example:

For the first row of df1 % df2:

0 December 285 of df1 matches with 0 December 285 of df2 so the dictionary will be

{285: 67511}

For the second row of df1 % df2:

1 December 296 of df1 matches with 1 December 296 of df2 so the dictionary will be

{285: 67511, 296:67512}

The first and second rows go together as they have the same index value (i.e December).

And so on for the other rows

Any ideas

CodePudding user response：

Let us merge the dataframes then group the resulting dataframe by index and create the record for each unique index inside a dict comprehension

{k: dict(zip(g['index1'], g['eventid'])) 
    for k, g in df2.merge(df1).groupby('index')}

{'December': {'285': '67511', '296': '67512'},
 'November': {'397': '64515', '300': '64517'}}