I have
import pandas as pd
df = pd.DataFrame({"A":[1,5,2,7,8],
"B":[5,1,7,2,9]})
# A B
# 1 5
# 5 1
# 2 7
# 7 2
# 8 9
1-5 should be considered as a duplicate of 5-1, and 2-7 to 7-2 and dropped to create:
# A B
# 1 5
# 2 7
# 8 9
The way I'm solving it now is to create a list of A and B, sort it and convert to string, then drop duplicates. There must be some easier and more effectivate way?
df["C"]=df[["A","B"]].values.tolist()
df["C"]=df.apply(lambda x: ','.join([str(y) for y in sorted(x["C"])]), axis=1)
# A B C
# 0 1 5 1,5
# 1 5 1 1,5
# 2 2 7 2,7
# 3 7 2 2,7
# 4 8 9 8,9
df = df.drop_duplicates(subset="C")
# A B C
# 0 1 5 1,5
# 2 2 7 2,7
# 4 8 9 8,9
