1

Can anyone tell me why df.loc can't seem to work in a loop like so


example_data = {
    'ID': [1,2,3,4,5,6],
    'score': [10,20,30,40,50,60]
}
example_data_df = pd.DataFrame(example_data)

for row in example_data_df:
    print(example_data_df.loc[row,'ID'])

and is raising the error "KeyError: 'ID'"?

Outside of the loop, this works fine:

row = 1
print(example_data_df.loc[row,'ID']

I have been trying different version of this such as example_data_df['ID'].loc[row] and tried to see if the problem is with the type of object that is in the columns, but nothing worked.

Thank you in advance!

EDIT: If it plays a role, here is why I think I need to use the loop: I have two dataframes A and B, and need to append certain columns from B to A - however only for those rows where A and B have a matching value in a particular column. B is longer than A, not all rows in A are contained in B. I don't know how this would be possible without looping, that would be another question I might ask separately

user20501139
  • 51
  • 7
  • In your own words, where the code says `for row in example_data_df:`, what do you think that will make `row` equal to each time through the loop? (Did you **check that**?) Now, does it make sense to use that result as an index for `.loc`? – Karl Knechtel Nov 14 '22 at 19:40

1 Answers1

0

If you check 'row' as each step, you'll notice that iterating directly over a DataFrame yields the column names.

You want:

for idx, row in example_data_df.iterrows():
    print(example_data_df.loc[idx,'ID'])

Or, better:

for idx, row in example_data_df.iterrows():
    print(row['ID'])

Now, I don't know why you want to iterate manually over the rows, but know that this should be limited to small datasets as it's the least efficient method of working with a DataFrame.

mozway
  • 194,879
  • 13
  • 39
  • 75
  • Thank you, it works! What I'm trying to do is: I have two dataframes A and B, and need to append certain columns from B to A - however only for those rows where A and B have a matching value in a particular column. B is longer than A, not all rows in A are contained in B. I don't know how this would be possible without looping, that would be another question I might ask separately – user20501139 Nov 14 '22 at 19:36
  • Yes better ask a separate question, make sure to provide a reproducible example of the DataFrames – mozway Nov 14 '22 at 19:49