Sub setting the dataframe and renaming based on the value of subset

Asked Feb 21 '19 at 12:33

Active Feb 21 '19 at 13:16

Viewed 42 times

I have a dataframe which looks like

df = pd.DataFrame({'Fruits':['Apple','Avacado','Banana','Bluberry'],'Alphabet':['A','A','B','B']})

I want to subset this data frame based on the alphabet and name the subsets with the corresponding alphabet

The straight forward way of doing this is

A = df.loc[df['Alphabet'] == 'A']
B = df.loc[df['Alphabet'] == 'B']

But I would like to automate it, if in case I have all the 26 alphabets in the dataframe.

I tried putting the unique values of alphabets in to a list

alphabets = list(df['Alphabet'].unique())

And then created a list of individual dataframes

dfs = []
for letters in alphabets:
    df = df.loc[df['Alphabet']==letters]
    dfs.append(df)

Is it possible to assign the item names of the list alphabets to dataframe items of list dfs

asked Feb 21 '19 at 12:33

Ahamed Moosa

I suggest create dictionary by this [solution](https://stackoverflow.com/a/45400233/2901002) – jezrael Feb 21 '19 at 12:36
@jezrael yes its an elegant solution and I have tried that in a different way too, but how to rename that dictionary key name to 'a' (corresponding alphabet) and store it as a data frame? – Ahamed Moosa Feb 21 '19 at 12:40
It is not necessary, because for keys in dictioanry ket alphabets and for values get DataFrames. – jezrael Feb 21 '19 at 12:41
1

But it is possible, by [globals](https://stackoverflow.com/a/45567689/2901002) – jezrael Feb 21 '19 at 12:42
Yeah , I get that. But ultimately I want to store the subsets as dataframe – Ahamed Moosa Feb 21 '19 at 12:43
1

@jezrael Thank You , I was looking for this globals() – Ahamed Moosa Feb 21 '19 at 13:14

0 Answers0