[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Column labels with --full
From: |
Erik Auerswald |
Subject: |
Re: Column labels with --full |
Date: |
Wed, 18 May 2022 09:00:01 +0200 |
Hi,
On Sun, May 15, 2022 at 06:07:56PM -0700, Dima Kogan wrote:
> As you say, this option isn't well-defined, and it doesn't make a lot of
> sense with aggregating operations. I think it makes great sense to
> disallow it in those contexts and throw an error, but that might break
> somebody's usage.
>
> Since it isn't well-defined and weird, I think we can decide what makes
> the most sense. I'm thinking of -f as "include all the other columns",
> and with that interpretation, it shouln't re-label anything that would
> be included without -f (i.e. the "GroupBy" in the field label). Also,
> with -g, it's important to have the "GroupBy" in the label, I think.
>
> Anybody else have opinions about it?
I concur that --full should not remove the GroupBy() part of the label,
or possible other label modifications when no --full is given.
Other than that I do not have a strong opinion regarding this.
Thanks,
Erik