Extract Specific Rows and Columns from a DF
Question:
I have the below data frame and I need to figure out how I can filter and only extract the info I need. The DF has groups and users. The A in the DF means the user is a group admin and U means just a normal user in the group.
I need to be able to get a report that shows me if any of the Groups listed in the DF has no admins(A), another report that gives me a list of Groups with no Admins or Users and finally a list of any users that are not members of any groups.
Name
Group1
Group2
Group3
Group4
Group5
Group6
User1
A
U
A
User2
A
A
U
User3
User4
A
A
U
User5
A
User6
A
I basically need 3 separate reports to come out of this 1 DF.
I have been trying to use Pandas, melt, dropna and groupby.agg but I am struggling to get the desired outputs.
Answers:
You can set_index on Name, then check boolean conditions aggregated with any/all for boolean slicing on the indexes:
df2 = df.set_index('Name')
# groups with no admins
df2.columns[~df2.eq('A').any()]
# or
# df2.columns[df2.ne('A').all()]
# ['Group4', 'Group6']
# groups with no user nor admin
df2.columns[df2.isna().all()]
# ['Group4']
# users part of no group
df2.index[df2.isna().all()]
# ['User4']
NB. I am assuming that empty cells have NaNs, if they have empty strings, replace isna() by eq('').
I have the below data frame and I need to figure out how I can filter and only extract the info I need. The DF has groups and users. The A in the DF means the user is a group admin and U means just a normal user in the group.
I need to be able to get a report that shows me if any of the Groups listed in the DF has no admins(A), another report that gives me a list of Groups with no Admins or Users and finally a list of any users that are not members of any groups.
| Name | Group1 | Group2 | Group3 | Group4 | Group5 | Group6 |
|---|---|---|---|---|---|---|
| User1 | A | U | A | |||
| User2 | A | A | U | |||
| User3 | ||||||
| User4 | A | A | U | |||
| User5 | A | |||||
| User6 | A |
I basically need 3 separate reports to come out of this 1 DF.
I have been trying to use Pandas, melt, dropna and groupby.agg but I am struggling to get the desired outputs.
You can set_index on Name, then check boolean conditions aggregated with any/all for boolean slicing on the indexes:
df2 = df.set_index('Name')
# groups with no admins
df2.columns[~df2.eq('A').any()]
# or
# df2.columns[df2.ne('A').all()]
# ['Group4', 'Group6']
# groups with no user nor admin
df2.columns[df2.isna().all()]
# ['Group4']
# users part of no group
df2.index[df2.isna().all()]
# ['User4']
NB. I am assuming that empty cells have NaNs, if they have empty strings, replace isna() by eq('').