Find a column1 value that has transacted in all distinct column2 values-CodePudding

I have to find all values in column1 that have transacted in all distinct values in column2. For ex:

From the above table we can identify the Fruitseller 'A' sells all the fruits.

CodePudding user response：

You can reshape values by crosstab first and then filter all index values if not 0 by boolean indexing:

df1 = pd.crosstab(df['Fruitsellers'], df['Fruits'])

out = df1.index[df1.ne(0).all(axis=1)].tolist()
print (out)
['A']

CodePudding user response：

>>> df
   Fruitsellers     Fruits
0             A      Guava
1             A      Mango
2             B      Apple
3             C     Banana
4             D  Pineapple
5             A     Orange
6             B     Orange
7             C  Pineapple
8             A      Apple
9             A     Banana
10            A  Pineapple

>>> filetr_func = lambda x:sorted(x.Fruits.unique()) == sorted(df.Fruits.unique())
>>> (df
...  .groupby("Fruitsellers")
...  .apply(filetr_func)
...  .where(lambda x:x==True)
...  .dropna()
...  .index
...  .to_list()
... )

['A']

CodePudding user response：

Try:

>>> df.drop_duplicates().value_counts('Fruitsellers').eq(df['Fruits'].nunique()) \
      .loc[lambda x: x].index.tolist()

# Output:
['A']

Assume in the Fruits column at least one instance is present.

drop_duplicates is not mandatory if you are sure, you have not multiple same rows of (Fruitsellers, Fruits) like 2 x ('A', 'Mango').