Map DataFrame column to 2 different dictionaries

Provided is code for how much allowance a kid would get based on country and age.
I can get the right allowance for a kid using a for loop but would like to
just use the map function instead.
If the country feature wasn't there I would just have been able to map
a kid's age value directly to 1 dict. The problem now is that I have 2 dicts I want
to map df['age'] to. Is it possible to get the right allowance using
only the map function and avoiding the for loop?

import pandas as pd

d1 = {
    'canada': {
        '0-5': 5,
        '6-8': 8,
        '9-10': 9,
        '11-15': 14,
        '16-18': 24},
    'us': {
        '0-5': 7,
        '6-8': 9,
        '9-10': 10,
        '11-15': 15,
        '16-18': 35}}

df = pd.DataFrame.from_dict({
    'name': ['Askeladd', 'Torkel'],
    'country': ['canada', 'us'],
    'age': ['0-5', '16-18']})

df2 = df.copy()
df2['allowance'] = [0]*2

# Gives right answer
for i in range(len(df2)):
    df2['allowance'].iloc[i] = d1[df2.iloc[i,:].loc['country']][df2.iloc[i,:].loc['age']]

# Gives wrong answer
df['allowance'] = df['age'].map(d1['us'])

Join

First make a Series out of d1:

s1 = (pd.DataFrame(d1).rename_axis(index='age', columns='country')
      .unstack().rename('allowance'))

• Or a more manual approach:

  d1_flat = {(co, age): x for co,a in d1.items() for age,x in a.items()}
  s1 = pd.Series(d1_flat, name='allowance').rename_axis(['country', 'age'])
  

Then join:

df.join(s1, on=['country', 'age'])

       name country    age  allowance
0  Askeladd  canada    0-5          5
1    Torkel      us  16-18         35

You can create a dataframe from your dictionary then merge with original:

df.merge(pd.DataFrame(d1).stack().rename('allowance'),
         left_on=['age','country'], right_index=True, how='left')
)

Output:

       name country    age  allowance
0  Askeladd  canada    0-5          5
1    Torkel      us  16-18         35