chaining and "filtering" datasets within MODIS h4 data

[lwasser]

Hi again @snowman2 ,
The clip approaches that we've discussed have worked well for all data with the exception of MODIS h4 data. that structure I believe has a main dataset and then subdatasets. When I use rxr.open_raster("path-here.h4") it returns a list of two xarray objects, a qa object and an object with the spectral "band" data in it.

In this instance where the hierarchical data returns two objects, is there a clever way to clip the output data that takes advantage of from_disk (e.g.rxr.open_raster("path-here.h4").clip(geometry, from_disk=True) to reduce memory and resource load? Right now we have to take each object and clip individually as an error is returned when I attempt to chain functions. Many thanks!

@nkorinek just pinging you here too so you can watch the thread too!

[snowman2]

You can use the variable/group kwargs to filter. See tests for examples:

rioxarray/test/integration/test_integration__io.py

Line 211 in 22251db

def test_open_group_filter():

[lwasser]

that did it! thank you. @snowman2 where did you get the group name from (MODIS_Grid_500m_2D)?

I also noticed that i can just grab the specific bands that i want using variable= in the tests. That also works great.
Thank you. that will save a lot of processing time.

[snowman2]

where did you get the group name

You can get it from inspecting the file (gdalinfo or by inspecting the subdatasets attribute on a rasterio dataset).