Problem with generate_mask.py

[HN-Jiang]

Dear all,

I am working on a study area with low coherence. In the invert_network step, I set mintpy.networkInversion.minTempCoh to 0.25. However, when I use view.py to check the resulting maskTempCoh.h5, it shows all values as 1 (no pixels are masked).

I have tried using generate_mask.py to manually adjust the threshold, but regardless of what I set the -m value to (e.g., 0.01 or 0.4), the output maskTempCoh.h5 remains entirely green (all values = 1). Could you please help me identify what I might be doing wrong?

Thank you for your help. Since GitHub does not support uploading .h5 files, I have uploaded the relevant data to Google Drive here

https://drive.google.com/drive/folders/1UTLs7potViJvi7nkImLovmiIWiqKznYS?usp=drive_link

https://drive.google.com/drive/folders/1UTLs7potViJvi7nkImLovmiIWiqKznYS?usp=drive_link

• view.py temporalCoherence.h5

• generate_mask.py avgSpatialCoh.h5 -m 0.2 -o maskTempCoh.h5
• view.py maskTempCoh.h5

[HN-Jiang]

I am not entirely sure, but this seems related to the Data Type of the mask file. I tried writing a generate_mask script and found that if I save it as a bool type, view.py displays all values as 1; however, after converting it to int8 (or uint8), the mask area displays and loads correctly.

By the way, if I understand correct, maskTempCoh.h5 should cover my research area when I use Deramp right?

[yunjunz]

By the way, if I understand correct, maskTempCoh.h5 should cover my research area when I use Deramp right?

Yes