StaMPS step error

Hi,

when I run step 2 of StaMPS, I encountered the following problem:


I don’t know why?
I need your help!
Thanks a lot!

Yours,
Dingfeng Duan

Please take a look at the following post, the solution is available, Also please try up to continue there,

Stamps (2,2) problem,

Source of the post

It works now!
Thank you very much!

Hi dear friends
I have a problem in StaMPS processing with MATLAB. I have faced with this error in STEP 1. Could anyone help me to fix this problem, please?
Thank you so much

How much patches did you create in total? 17 seems a bit large to me.

I created 459 patches

way too much, if you ask me, especially considering the selected overlap.

What exact command did you enter when calling mt_prep?

I enter this command: mt_prep_snap “master date” / “INSAR_master date’s directory”/ 0.4 27 17

before I did StaMPS processing with less patches for another data series, but encountered with memory error in STEP 5, so I increased patches and StaMPS process done successfully. my study area is too wide.

no option to reduce the size of the study area or to move to a stronger PC for this task?
I just think that the tiles will be so small that their size is conflicting with the overlap which is required to merge them later.
Especially because this happened for PATCH 17 indicates that it is related to a tile at the edge which is probably too small in the end.

Have you tried a different combination? e.g. 0.4 21 21

unfortunately, It is not possible to reduce the size of area and we have not any other PC.
I start to do process with this combination you said now.
Thank you so much

Hi dear friend
I tried this combination 0.4 21 21 and encountered with the same error but this time in patch 20

now it happened for patch 20 (the last in line) which is again too small - at least this is my guess.
I don’t see a solution without larger tiles, sorry.

Thank you so much. I try with new combination e.g. 0.4 19 14. I hope this work

Have u finished it? I am encountering the same problem.

Hello everyone! I am facing a similar problem processing 2 entire sub-swaths. I had increased my ram to 64 gb but can’t complete step 2 with 15 images. Wich variable allows me to use at least 2 patches when processing? Because when i prepare the data i only enter 0.4 in the code line at the end. Should I use 0.4 2 1 to obtain a total of 2*1 patches?
Thanks a lot!

Yes, first you define the amplitude dispersion, then the number of rows and colums and then the overlap: 0.4 2 1 50 200

That can be a game changer for me. I will try it out rn. Thanks a lot!

I got this message when after 20 hours of processing.


Any ideas of how to solve it?
Thanks!

Hello everyone!
Yesterday I figured out how to solve this when I was processing other AOI. You have to change the script ps_correct_phase.m in lines 44 and 49 to add the additional parameter ‘-v7.3’ to the save function. With this parameter Matlab is forced to use version 7.3 to handle variables bigger than 2 Gb. That change allowed me to complete step 5 without errors and to finish successfully the whole process.
This error occurs because without using version 7.3 the variable ph_rc is not stored and therefore not found later. -v7.3 parameter is slower so maybe that is why the developer didn’t include it as default. Other scripts also uses the save function without contemplating this so this change may apply to other similar errors.
ps_correct_phase.m script before the change:
44- save(rcname,‘ph_rc’);
45- …
46- …
47- …
48- …
49- save(rcname,‘ph_rc’,‘ph_reref’);
ps_correct_phase.m script after the change:
44- save(rcname,‘ph_rc’, ‘-v7.3’);
45- …
46- …
47- …
48- …
49- save(rcname,‘ph_rc’,‘ph_reref’, ‘-v7.3’);

thank you for reporting. Can you please confirm the error message which is solved by your solution?