Pixel weighting option about mintpy HOT 3 CLOSED

Hi Rob,

I am not sure if I got your interest completely.

The phase variance (var or fim) is converted from filtered coherence.

To assign a low weight to pixels with connected component of 0 is equivalent to mask out those pixels. Have you tried the mintpy.networkInversion.maskDataset option? This option helps to increase the spatial coverage (temporal coherence) when the kept interferograms (with overall moderate or high coherence) has some low coherent pixels. This option will result in pixels with different numbers of interferograms, as explained in section 4.5.1 in our paper.

We could meet on the Caltech campus if that's easier to discuss.

from mintpy.

Hi Yunjun,
Ah, yes! I think the maskDataset option is what I am looking for... As I understand, instead of masking out a pixel entirely based on the connComp mask, it only masks out interferograms when which the connected component is zero. Thus, the pixels can contain different numbers of interferograms. Sorry, I'd missed that option last time I'd read through.

It would be great to meet up at Caltech one day.

from mintpy.

I am glad that you got what you are looking for!

from mintpy.