CODEWITHSUNDEEP
python-3.xdictionary
Ganitph
Ganitph

Reputation: 123

How to dynamically create a dictionary from product of 2 lists at a time

I have a large number of iterations (~60000) on each I get 2 lists at a time.
I need to keep some information about each possible combination of items from first/second list. if a pair is repeated I want to count it - not add, that's why I think dictionaries are more suitable.
I can do it with nested loops - but it takes forever.
I thought of using a dictionary of dictionaries.
I can tell all possible values ahead so I started with a dictionary with empty values - I want to add them iteratively.

If this is the input on the first iteration

l1 = ['1','2']
l2 = ["A", "B"]

I expect to get:

{'1':{'A': [1, 0], 
      'B': [1, 0]}
 '2':{'A': [1, 0], 
      'B': [1, 0]},
 '3': {},
 '4': {},
 '5': {}}
 }

and after a second iteration with: 

l1 = ['1','3','5']
l2 = ['B','C']]

I expect to get:

{'1':{'A': [1, 0], 
      'B': [2, 0],
      'C': [1, 0]}
 '2':{'A': [1, 0], 
      'B': [1, 0]}
 '3':{'B': [1, 0], 
      'C': [1, 0]},
 '4':{},
 '5':{'B': [1, 0], 
      'C': [1, 0]}
 }

What's the most efficient way of doing it?

Upvotes: 0

Views: 31

Answers (1)

Venkatachalam
Venkatachalam

Reputation: 16966

If your lists are in large size, then I would suggest linked list sparse matrix.

Try this!

l1_superset = ['1','2','3','4','5']
l2_superset = ['A','B','C','D','E']

M=sparse.lil_matrix(np.zeros((len(l1_superset),len(l2_superset))))

mapper_l1 = {k:idx for idx,k in enumerate(l1_superset)}
mapper_l2 = {k:idx for idx,k in enumerate(l2_superset)}

def update (l1,l2):
    for i in l1:
        for j in l2:
            M[mapper_l1[i],mapper_l2[j]] = M[mapper_l1[i],mapper_l2[j]] + 1

l1 = ['1','2']
l2 = ["A", "B"]
update(l1,l2)

l1 = ['1','3','5']
l2 = ['B','C']
update(l1,l2)

Upvotes: 1

Related Questions