filter a Pandas dataframe for added unique values

Question

I would like to know what I need to do in order to filter a dataframe, keeping unique values of Name column, adding values from Value column and adding a new column for counting appearances of each Name

what I have is this:

     Name Type  Value
0   apple    A      1
1  banana    B      3
2   apple    A      2
3    pear    P      4
4   apple    A      6
5  carrot    C      3
6  banana    B      2

and I want to filter it into this:

     Name Type  AddedValue  Occurrences
0   apple    A      9       3
1  banana    B      5       2
2    pear    P      4       1
3  carrot    C      3       1

How can I do it? I've tried conceive a .join method with a where condition set, but I cannot make it work and I'm sure that the problem is that I'm trying to translate pythonic thinking where there surely is a panda instruction which solves my problem with an elegant vector operation or something like that

Thanks in advance

Answer 1

Try groupby method:

df.groupby(["Name","Type"]).agg(["count","sum"])

Result:

            Value    
            count sum
Name   Type          
apple  A        3   9
banana B        2   5
carrot C        1   3
pear   P        1   4

However if You want to flatten columns/index use:

df2 = df.groupby(["Name","Type"]).agg(["count","sum"]).reset_index(drop=False)

df2.columns = [' '.join(col).strip() for col in df2.columns.values]

Output:

     Name Type  Value count  Value sum
0   apple    A            3          9
1  banana    B            2          5
2  carrot    C            1          3
3    pear    P            1          4

Even more elegant solution thanks to @piRSquared:

df2 = df.groupby(['Name', 'Type']).Value.agg([('AddedValue', 'sum'), ('Occurences', 'count')]).reset_index(drop=False)

Output:

     Name Type  AddedValue  Occurences
0   apple    A           9           3
1  banana    B           5           2
2  carrot    C           3           1
3    pear    P           4           1

Answer 2

Yes, just like ipj has answered you can try the groupby method in Pandas Groupby .

df = df.groupby(["Name","Type"]).agg(["count","sum"])
df.columns = df.columns.droplevel(0)
df = df.rename({"count": "AddedValue", "sum": "Occurrences"}, axis=1)