Specifying a range of columns as parameter

Question

I want to extract the max value of a table that has 100 rows: Something that can be poorly achieved like this:

maxA = max(data['a1size'].max(), data['a2size'].max(), data['a3size'].max()... and continuing like this until a100size.

Of course that solution will be not ideal at all.

Is there a better approach and more elegant than using a for loop?

I'm looking for some kind of .. operator. Something like:

maxA = max(data['a1size'].max() .. data['a100size'].max())

Answer 1

The loc operator allows you to use slicing by column names:

maxA = data.loc[:, 'a1size':'a100size'].values.max()

Note that in contrast to core Python slicing, the final column is included here. The first : specifies that you want all the rows.

This works if the columns you want to take the maximum over are contiguous. Otherwise, you could index the dataframe with a list of column names instead. To construct such a list, you could use a list comprehension:

a_size_cols = [f'a{i}size' for i in range(1, 101)]

maxA = data[a_size_cols].values.max()