wrong dtype in MSTP example (rvt_vis_example.ipynb)

[nanttiroiko]

MSTP example in examples/rvt_vis_example.ipynb states that the output is of int8 dtype, but actually it is float32. Other documentation and code consistently point to intended dtype being float32.

[zkokalj]

This needs further exploration. Output I get processing a single tile is 8-bit.

[nanttiroiko]

Curious - but also possible that I'm just getting something wrong.

Here is a minimal code example I can use to print out mstp_arr dtype as float32

import rvt.vis

dem_path='dem\\N5421C4_1_DEM.tif'

dict_dem, dem_arr, dem_resolution, dem_res_x, dem_res_y, dem_no_data = read_dem(dem_path)

mstp_arr = rvt.vis.mstp(dem=dem_arr, local_scale=(1,5,1), meso_scale=(5,50,5),broad_scale=(50,500,50), lightness=1.2, ve_factor=1, no_data=dem_no_data)

print(mstp_arr[0].dtype)

In case it matters, I'm running the following version of rvt-py:
rvt-py 2.2.1 pypi_0 pypi

[zkokalj]

In this case you'll get a float result. You are correct, I was looking at a wrong Jupyter Notebook. MSTP in rvt_vis_example.ipynb normalises to 0-255, but writes a float32 dtype. It should be corrected to output the 8-bit file.

[nanttiroiko]

Great. Just to clarify I'm happy with how rvt.vis.mstp behaves and only found the example confusing.