[FIXED] Multiindex dictionary - transform dict with other indexes - look up dictionary

Issue

I am trying to transform my dictionary to a "look up dictionary" with multiindexes.

I have the following dictionary where the current keys are the period:

{8: {'FlightName': ['Flightname612'],
  'Flight_idx': array([268]),
  'Baggage': array([149]),
  'infeed_idx': [1]},
 9: {'FlightName': ['Flightname612'],
  'Flight_idx': array([268]),
  'Baggage': array([149]),
  'infeed_idx': [2]},
 10: {'FlightName': ['Flightname353', 'Flightname394', 'Flightname612'],
  'Flight_idx': array([  9,  50, 268]),
  'Baggage': array([ 21,   1, 149]),
  'infeed_idx': [1, 1, 3]},

And I would like a dictionary where I can look up the flight index and the infeed index and it will then return then period. So the expected out:

(flight_idx, infeed_idx)=period
(268,1) = 8
(268,2) = 9
(9,1) = 10
(50,1) = 10
(268,3)=10

I was out trying to solve it using a loooot of loops, but I feel like there is a simple solution, if you know your way around pandas, dictionaries and operations. I hope you can help!

Solution

You can create a dataframe with data as your input dictionary's values and index as its keys.

Then reset_index() to make a new column for period and set_index() with [Flight_idx, infeed_idx] and use that to build a dictionary by using to_dict.

Code:

df = pd.DataFrame(in_dict.values(), index=in_dict.keys())
out_dict = df.explode(["Flight_idx", "infeed_idx"]).rename_axis("period").reset_index().set_index(
    ["Flight_idx", "infeed_idx"]
)["period"].to_dict()

print(out_dict):

{(268, 1): 8, (268, 2): 9, (9, 1): 10, (50, 1): 10, (268, 3): 10}

Answered By - SomeDude