On Wed, 22 May 2019, Riccardo (Jack) Lucchetti wrote:
On Wed, 22 May 2019, Sven Schreiber wrote:
> Hi,
>
> I'm using the 'join' command on a lot of source files with semicolon as
> the field separator. This seems to work, but I was wondering whether it
> would accelerate the whole thing if I could tell gretl about the
> semicolon? I haven't found an option for this in the user guide chapter
> on 'join'.
>
> (It could also be that it is just slow because it's a lot of files...)
Of course I'm not in the position of making an informed judgement, but I
reckon that the CPU time needed to figure out what the delimiter is should be
2 or 3 orders of magnitude smaller than anything involving I/O...
True, but it could be worth trying
set csv_delim semicolon
This is documented as working on the output side, but I believe it
also gives a hint on the input side.
Allin