Thanks John, Markus ! I'll use that !

On Tue, 14 Dec 2021 at 11:41, John Darrington <j...@darrington.wattle.id.au>
wrote:

> I concur with Markus.  Probably the easiest way to do what you want
> is with a command similar to:
>
> pspp regression.sps | awk -F '|' '/Step 1 depth/{print $7}'
>
> J'
>
>
> On Tue, Dec 14, 2021 at 07:36:05AM +0000, Quandt, Markus wrote:
>      Hi,
>
>      I don???t think you can do this in PSPP syntax (but would be happy to
> be corrected).
>
>      But since you are working on a Unix/Linux system anyway, why not ask
> someone who is familiar with regex (regular expression) to parse the output
> file and write out the ???hits??? to a separate csv file? Sorry, I cannot
> help with the specific regex syntax, as I am not a Unix person.
>
>      Best,
>      Markus
>
>      Von: Pspp-users <pspp-users-bounces+markus.quandt=gesis....@gnu.org>
> Im Auftrag von Oren Ish-Shalom
>      Gesendet: Dienstag, 14. Dezember 2021 06:12
>      An: pspp-users@gnu.org
>      Betreff: Output ONLY sig. level in binary logistic regression
>
>
>      Hi all,
>      I???m running 200,000 binary logistic regressions
>      (loop from a bash linux terminal)
>      I???m only interested in extracting the significance level.
>      Here is my sps script:
>
>      GET DATA
>
>      /TYPE=TXT
>
>      /FILE='example_data.csv'
>
>      /DELIMITERS=','
>
>      /FIRSTCASE=2
>
>      /VARIABLES=
>
>          serial F32
>
>          gene A64
>
>          time F8.8
>
>          survived F1.
>
>
>
>      LOGISTIC REGRESSION survived WITH time.
>
>      EXECUTE.
>
>      When I run it I get three tables but only need the significance
> (0.436 here):
>
>                  Variables in the Equation
>
>      +---------------+-----+----+----+--+----+------+
>
>      |               |  B  |S.E.|Wald|df|Sig.|Exp(B)|
>
>      +---------------+-----+----+----+--+----+------+
>
>      |Step 1 depth   | 2.36|3.03| .61| 1|.436| 10.62|
>
>      |       Constant|-2.75|2.90| .90| 1|.344|   .06|
>
>      +---------------+-----+----+----+--+----+------+
>
>      Any way to print just this value ?
>      Thanks !
>

Reply via email to