.normalize() doesn't work with .where()

[critmcdonald]

I have a case where I can't filter a table after it has been run through .normalize(), getting an error IndexError: tuple index out of range.

It is a repeatable problem, but it is not always a problem. I've proven that I can also filter a simple normalized table. So, while this could be user error, I don't think so, but I would be happy to be proven wrong.

I have a repo that reproduces the problem in a notebook. But I also have an example where I can filter a normalized table.

[gutte]

I also get IndexError: tuple index out of range when performing certain operations on a table that has been normalized using .normalize(). It seems to be something about how the column names are indexed.

Example:
table = agate.Table.from_csv('testinput.csv')
norm_table = table.normalize('student',['course1','course2','course3','course4'])
selection = norm_table.where(lambda row: row['property'] == 'course1') # THIS GIVES AN ERROR

where testinput.csv:
student,course1,course2,course3,course4
A,1,2,3,4
B,3,4,1,2
C,2,1,3,4
D,1,4,3,2
E,3,4,1,2

My reason for thinking that its not a user issue is that the problem can be worked around by temporarily writing the table to file and reading again, hence "fixing" the indexing:

norm_table.to_csv('temp.csv')
norm_table = agate.Table.from_csv('temp.csv')
selection = norm_table.where(lambda row: row['property'] == 'course1') # NO ERROR!

[jpmckinney]

I think the issue is that normalize adds new rows, but it doesn't fill in row names for the new rows, so you end up getting an index error.

[jpmckinney]

@gutte Thank you for sharing the easy-to-replicate code!