[Gretl-users] Re: Gretl: means by categorical variable
Am 26.05.2024 um 23:32 schrieb Cottrell, Allin:
On Sun, May 26, 2024 at 8:43 AM Sven Schreiber
<sven.schreiber(a)fu-berlin.de> wrote:
> Thanks, Allin, I can see the result table. However, the numbers are
> flowing into each other, I think that separating spaces are needed
> between the columns. Example:
>
> AreaRegion Mittel min Q1 Median
> Q3 max
> Australia - Oceania 1.5576e+006 1647.0
> 14750.1.0373e+0052.9862e+0052.5810e+007 (n=22)
Oof, that's a Windows-specific problem. (The Microsoft C library
insists on printing the exponent of a floating point number to 3
digits, which for socioeconomic data generally means with two leading
zeros.) Maybe we should try skipping the extra leading zero in our
Windows builds.
I have also noticed the e+00x (three digits) format in the context of
the summary command output on Windows, which struck me as mostly
unnecessary and distracting. Is that different on Linux as well?
Related to this kind of formatting, was my hypothesis correct that the
5-significant-digit limitation for the summary command was chosen
because of the field width? Otherwise it's not so obvious why the max
for the 'display_digits' setting is 6 in general but only 5 for summary.
Perhaps this could be harmonized if we can save one of the digits in the
e+00x part.
cheers
sven
Attachments:
- attachment.htm (text/html — 2.2 KB)