Multi-mode inversion with Dinver from command prompt

[yaolinm]

Dear Dr. Wathelet,

Thank you for the super helpful tool of Dinver and I sincerely appreciate your time for maintaining this forum and answering questions. I benefit a lot from previous posts and replies and I am able to understand most of the usages of Dinver. My question is actually a follow-up from the previous post of multi-mode inversion.

While I understand I could do multi-mode inversion in Dinver GUI, I am wondering if it's possible to use the standalone executable of Dinver to run multi-mode inversion from terminal? I am trying to do multi-mode inversion for a very dense array, and it would save me a lot of time to run from terminal with a loop than manually input files from the GUI. I saw the documentation of terminal usage and the input file for target seems to be limited to one file. Is there a way to specify multiple targets for multiple mode in terminal?

Thank you again for your time!
Yaolin

[admin]

Hi Yaolin,

Sorry for the very late answer. While sorting emails from last year, I found a lot of pending requests for geopsy forum.

The command line interface can be used also for multi-mode. The multi-mode is defined in the target file. If I'm right, gptarget can aggregate several curves into a single target file. Each curve can have a different mode index. From the help message, it is probably not possible to set several modes for a curve like in the graphical user interface. The target file is a compressed xml file. An alternative is to build the xml contents in a script.

Setting several modes for a curve should be added to gptarget. I'll check it.

Best regards,

Marc

[zibi]

Hi Yaolin and Marc,

I am using a multi-mode inversion from a command line. I will drop the command lines here, maybe you will find them useful.

Code: Select all

	
cat DC1.disp | gptarget A -dispersion-rayleigh 0 target1.target
cat DC2.disp | gptarget A -dispersion-rayleigh 1 target1.target

dinver -i DispersionCurve -optimization -target target1.target -param param_file.param -ns0 500 -ns 1000 -nr 50 -f -o output.report

However, it is not clear to me whether this way I am specifying the number of the mode or am I simply allowing the inversion to fit either fundamental mode or higher mode to DC2.disp. Marc, maybe you could cast some light on that?

I hope you find it helpful.
Zibi

[admin]

Hi Zibi,

To answer your question, open target1.target from the graphical interface. Menu File/Import targets. Curve DC1 is mode 0 and curve DC2 is mode 1. It is no OR.

You message suggested a modification of gptarget options. From version 3.6.0-preview (available from git repository), INDEX can be a list of integers separated by comas.

Code: Select all

cat DC3.disp | gptarget A -dispersion-rayleigh 2,3 target1.target

DC3 can be either mode 2 or mode 3.