CODEWITHSUNDEEP
pythonpandas
user1361529
user1361529

Reputation: 2697

Creating new unique rows from a list of possible column values

Lets assume I have a Pandas Dataframe like this:

   C1    C2    C3
0  1     B     v
1  5     D     i
2  1     B     iii
3  3     C     iv

Where all possible values of C1, C2, C3 are

C1 = [1,2,3,4,5]
C2 = ['A','B','C','D','E']
C3 = ['i','ii','iii','iv','v']

The ask is to print new rows to exhaust all possible combinations of C1, C2, C3 that are not already in the existing dataframe.

Is there a better way than to have 3 nested loops ranging all the values of C1, C2, C3 and comparing each combination to the existing list?

Upvotes: 0

Views: 49

Answers (2)

Jai
Jai

Reputation: 3310

series_1 = pd.Series(C1)
new_data_frame['C1'] = series_1.values
series_2 = pd.Series(C2)
new_data_frame['C2'] = series_2.values
series_3 = pd.Series(C3)
new_data_frame['C3'] = series_3.values
for index, row in new_data_frame.iterrows():
    if existing_dataframe[existing_dataframe == row].shape[0] == 0:
         existing_dataframe.appen(row)

Upvotes: 0

bphi
bphi

Reputation: 3185

You could try something like this,

C1 = [1, 2, 3, 4, 5]
C2 = ['A', 'B', 'C', 'D', 'E']
C3 = ['i', 'ii', 'iii', 'iv', 'v']
existing = ((1, 'B', 'v'), (5, 'D', 'i'), (1, 'B', 'iii'), (3, 'C', 'iv'))

import itertools
result = [i for i in itertools.product(C1, C2, C3) if i not in existing]

Upvotes: 1

Related Questions