I have some pandas Dataframes with repair order information, with each order containing several criteria c:
orders (o)
------
id c1 c2 c3 c4 c5
0 Car XL 4 Blue Metal
1 Jet XXL 0 Pink Metal
2 Car M 4 Blue Metal
3 Bike Med 2 Red Wood
There is also another list with repair plants. Each plant can have certain criteria c which makes them available only for orders with these criteria. No entry/NaN means that all orders are eligible for repair:
repair_plants (rp)
-------------
id Type c2 c3
0 Car NaN 4 --> Available for all Cars with 4 wheels
1 Car M 4 --> Available for all Cars of size M with 4 wheels
2 Jet NaN 0 --> Available for all Jets without wheels
rp is generated during runtime, it does not necessarily contain all criteria. A missing criteria should be handled as if it was NaN. I do not know beforehand how many colums with criteria the Repair Plants List has.
The goal is to find a list with all eligble repair plants per order. So for example for order 0 and 2, I would get:
Repair Plants for order 0 Repair Plants for order 2
------------- -------------
id Type c2 c3 id Type c2 c3
1 Car M 4 0 Car NaN 4
(0 is not eligble since c2="XL") 1 Car M 4
What is the best way to create these sublists? I know I can just use .loc and some comparisons if I know all the columns in rp beforehand, e.g.:
for i in range(len(o)):
value2 = o.loc[i,"c2"]
value3 = o.loc[i,"c3"]
print(rp[(rp.c2.isna()) | (rp.c2 == value2)) & ((rp.c3.isna()) | (rp.c3 == value3)])
But what do I do to adress the unknown number of criteria? Another loop inside my for loop? What would be the fastest and/or most elegant way of getting the list for each order?
