PSI using SNAP-STaMPS workflow in Windows Subsystem for Linux (WSL)

[suribabu]

Can you forward the error message with screen shot?

[mikhelif]

There’s no actual error message it just get stuck there

[suribabu]

Have you set the path of StaMPS-4.1-beta/bin in your PC.
May be below post is useful for you check it.

Workflow between SNAP and StaMPS

If you install all the software in root (using sudo), i think no need to add any path in Stamps_config.bash. only thing i added in .bashrc file stamps/bin like below open .bashrc file in root export PATH=path of stamps upto bin (/home/username/stamps/bin):$PATH then run source .bashrc And then run mt_prep_snap in any path, if it comes then are ready to use Stamps.

[suribabu]

Check this below link also.

New StaMPS release (4.1-beta)

mt_prep_snap has to be ran outside matlab (in the shell), it is a Linux script to preprocess the data exported by SNAP. If this was completed correctly, you proceed to open matlab in this folder and start the stamps scripts.

[mikhelif]

Thanks for your replies, I tried both both but unfortunately it still doesnt work.

Here’s a screenshot of what it looks like

[suribabu]

Check this below one. Best link given by ABraun.

Doris5.03 stamps stamps

I don’t know much about the data, so it is very hard to guess. What type of data do you use (sensor) and how large is your area? The number of PS candidates seems quite large to me. Also witdth = 1 and number of interferograms = 0 is very strange. Have you seen this documentation? https://gitlab.com/Rexthor/gis-blog/blob/master/StaMPS/2_stamps_workflow.md

[mikhelif]

After reading carefully everything on the screen, I think i figured out what may be causing it to freeze.
Here’s the first lines i get after executing mt_prep_snap:

Amplitude Dispersion Threshold: /0.4
Processing 1 patch(es) in range and 1 in azimuth

/home/ikhelif/Downloads/StaMPS-v4.1-beta/matlab/ps_parms_initial.m: No such file or directory.

I checked the folder and the file is actually there, still the same issue even after re installing StaMPS, Any ideas what I should try next ?

[gnwiii]

.m files are text, does less /home/ikhelif/Downloads/StaMPS-v4.1-beta/matlab/ps_parms_initial.m display the file? No such file or directory messages can occur when some directory along the path ended up using Unicode “look-alike” characters, for example, ASCII “minus” and “dash” are the same, while Unicode can have different dashes: “em-dash”, “en-dash”, “minus”, etc.

[mikhelif]

Thanks for your reply, I did rename the Stamps folder without the dash and I no longer get the “No such file or directory” error, yet still stuck at mt_extract_cands

[ABraun]

if I understood you correctly, you are not dividing your data into patches. Patch 1 (the only one existing) therefore needs considerable time for processing.
Please add 2 3 at the end of your mt_prep_snap command to divide your data into 6 patches (3 rows, 2 colums). This should speed up the processing a bit.

[mikhelif]

Thanks but I already tried that but it didn’t work, now with each patch i get this:

selpsc_patch /home/ikhelif/stamps/bin/selpsc.in patch.in pscands.1.ij pscands.1.da mean_amp.flt f 1
selpsc_patch: Command not found.

psclonlat /home/ikhelif/stamps/bin/psclonlat.in pscands.1.ij pscands.1.ll
psclonlat: Command not found.

pscdem /home/ikhelif/stamps/bin/pscdem.in pscands.1.ij pscands.1.hgt
pscdem: Command not found.

pscphase /home/ikhelif/stamps/bin/pscphase.in pscands.1.ij pscands.1.ph
pscphase: Command not found.

[ABraun]

this means stamps was not compiled correctly.

Did you run make and make install after downloading the package?

[mikhelif]

Yes pretty sure the installation went without problem, still can’t figure out why it’s not working

[panjibrotoisworo]

One issue I encountered with StaMPS is that you need gcc 7 and g++ 7. I have an Ubuntu 20.04 VM and the default version is gcc 9 and g++ 9. StaMPS failed to compile properly. I had to downgrade to version 7.

You can check your versions using this command:

[ABraun]

this is true. Please see here:

Segmentation Fault (core dumped) error during mt_prep_snap

@JeroBande is right, the problem is the compiler version. StaMPS has to be compiled with gcc-7 (gcc-9 was installed in my case and used for the compilation of StaMPS which led to incorrectly compiled scripts) So this is what it solved the problem in my case (partially copied from here). I cannot grant that it will solve all problems, but changing the compiler makes a difference. If you get stuck at one of these steps, please try to continue searching the web, because I have no clue how to troub…

[mikhelif]

I tried before, it fixed the “segmentation fault” problem but still it won’t run properly, i think i should re install Ubuntu/WSL

[gnwiii]

Reinstall should be a last resort, and often just gets you back in the same unhappy place.

Have you searched the forums for other reports of your current problem?

If you can provide more details of what doesn’t work someone may recognize the issue an suggest a fix. For starters, there is the original WSL and also WSL2 which uses a real linux kernel and a VM. If your system supports WSL2 and you are using the old WSL it is probably worth trying WSL2. The procedure to change uses wsl.exe and is
time-consuming. You can revert back to WSL if there are new problems with WSL2, but most people prefer WSL2. Also WSL2 will be getting native graphics.

Can you provide full details of your processing steps up to the one that doesn’t work so others can try to reproduce the issue? It is often easiest if you attempt to reproduce a tutorial as others will already have the tutorial data.

[suribabu]

As for my personal experience Ubuntu 18.04 LTS is best for StaMPS code. Try it once.

[mikhelif]

Thanks, currently switching to Wsl 2 and see what that does.

I tried reproducing the same steps in the tutorial above, but after launching mt_prep_snap it stops at mt_extract_cands

[mikhelif]

Update: after switching to Wsl 2 I’ve been having a “Segmentation violation detected” error message with matlab crashing right after that, I also found other crash reports saved from previous attemps meaning it’s the same issue even if it didn’t show up the same message.

Anyone knows how to fix it ?