Output from multiple runs in Dinver

This forum is dedicated to discuss all problems and suggestions related to the inversion software
Post Reply
Aniket19061990
Posts: 13
Joined: Sun Jan 24, 2021 11:11 am

Output from multiple runs in Dinver

Post by Aniket19061990 »

Hello.
I have 200 runs in a Dinver files, each with a different target but same parameterizations. I want -gpprofile of lowest misfit from each run. Currently I am giving -gpprofile command 200 times, each for a single run. Is there any shorter way for doing that?
admin
Site Admin
Posts: 841
Joined: Mon Aug 13, 2007 11:48 am
Location: ISTerre
Contact:

Re: Output from multiple runs in Dinver

Post by admin »

No, not really. If you merge all the best models produced by "gpdcreport -best 1" in a single stream, you can run gprofile only once

Code: Select all

(
  for F in $(ls *.report); do
    gpdcreport $F -best 1
  done
) | gprofile -vs
I'm unsure if this is what you meant.
Aniket19061990
Posts: 13
Joined: Sun Jan 24, 2021 11:11 am

Re: Output from multiple runs in Dinver

Post by Aniket19061990 »

Thanks for your reply.
I wrote the command as you have written, but it showed this error:

sh: syntax error near unexpected token `)'
admin
Site Admin
Posts: 841
Joined: Mon Aug 13, 2007 11:48 am
Location: ISTerre
Contact:

Re: Output from multiple runs in Dinver

Post by admin »

This scirpt is written in bash. If you save the code in a file with the first line set to

Code: Select all

#!/bin/bash
It should be better.
alcollins
Posts: 2
Joined: Thu Jan 19, 2023 10:07 pm

Re: Output from multiple runs in Dinver

Post by alcollins »

Hi Marc,

I have a similar use case to this so I thought it made sense to post in this thread rather than a new one.

I'm trying to generate a mean ellipticity curve and mean ground profile from three separate runs. I'd also like to be able to plot the 100 best misfits from each of the three separate runs on one plot. Using gpdcreport run-01.report run-02.report run-03.report -best 3 [or -best 300] won't work here because, for example, the best misfit (or best 100 misfits) from run 2 might be very different from the best misfit (or best 100 misfits) of run 1. I think what I'd like to do is combine all of these data into one report, which I can then manipulate and plot from as needed.

So I reviewed this thread, which seems to get at what I want. But when I try to execute a similar script to your example above using Git Bash on Windows, I get an error that the commands are not recognized. Is this an issue with my path? Could you help me understand a little better how Geopsy integrates with bash and how you'd run these scripts?

Here's the code I tried to use:

Code: Select all

#!/bin/bash

(
    for F in $(ls *.report); do
        gpdcreport $F -best 1
    done
) -o best_three.report
Cheers!
admin
Site Admin
Posts: 841
Joined: Mon Aug 13, 2007 11:48 am
Location: ISTerre
Contact:

Re: Output from multiple runs in Dinver

Post by admin »

Hi,

Replace "-o best_three.report" by " > best_three.model".
The output of gpdcreport is a text describing a 1D layered model, not binary data like in a .report.
Hope that it solve your issue

Marc
Post Reply