Union of all values on a conditional case-CodePudding

I am interested unpacking the values of dictionary that contains a list of values.

I need to combined all the values inside the dictionary for each key.

d1 = {
    'A': ['col1', 'col2'],
    'B': ['col3', 'col4'],
    'C': ['col5', 'col6']
}

The output I want is

d2 = {
    'A': ['col1', 'col2', '0 col3', '0 col4', '0 col5', '0 col6'],
    'B': ['0 col1', '0 col2', 'col3', 'col4', '0 col5', '0 col6'],
    'C' : ['0 col1', '0 col2', '0 col3', '0 col4', 'col5', 'col6']
}

d1 = {'A': ['col1', 'col2'], 'B': ['col3', 'col4'], 'C': ['col5', 'col6']}


c1 = [v for k, v in d1.items()]

d2 = {}

for k, v in d1.items():
    for l in c1:
        if l in v:
            d2[k] = l
        else:
            d2[k] = ','.join(l)

How can I unpack all the values for each key, combine them and a static value needs to be added for values not listed to the key.

CodePudding user response：

You're definitely on the right track.

d1 = {'A': ['col1', 'col2'], 'B': ['col3', 'col4'], 'C': ['col5', 'col6']}
all_values = [v for sublist in d1.values() for v in sublist]
d2 = {}

for key in d1.keys():
    new_values = []
    for v in all_values:
       if v in d1[key]:
          new_values.append(v)
       else:
          new_values.append('0 '   v)
    d2[key] = new_values

Output:

{'A': ['col1', 'col2', '0 col3', '0 col4', '0 col5', '0 col6'], 'B': ['0 col1', '0 col2', 'col3', 'col4', '0 col5', '0 col6'], 'C': ['0 col1', '0 col2', '0 col3', '0 col4', 'col5', 'col6']}

Shorter version:

d2 = {}

for key in d1.keys():
    d2[key] = [v if v in d1[key] else '0 '   v for v in all_values]

CodePudding user response：

Here are a couple of different approaches.

Make a second dictionary with the prefixed values:

d1b = {k: [f'0 {x}' for x in v] for k, v in d1.items()]}

Now make a list containing the two dictionaries for easier access:

ds = [d1b, d1]

You can now construct the output thus:

d2 = {k: sum((ds[k1 == k][k1] for k1 in d1), []) for k in d1}

This uses sum to apply the operator to all your values (notice the start value of []). The index k1 == k is a boolean which selects index 0 or 1 from ds.

Construct the sum of all the values:

values = [f'0 {e}' for v in d1.values() for e in v]

Also record the lengths their cumulative sum:

from itertools import accumulate

lens = [len(v) for v in d1.values()]
index = list(accumulate([0]   lens[:-1]))

Now you can do a selective replacement in values:

d2 = {}
for (k, v1), i, n in zip(d1.items(), index, lens):
   v2 = values.copy()
   v2[i:i   n] = v1
   d2[k] = v2