search

kmaterna / Strain_2D

Strain rate modeling from GNSS velocity fields
MIT License
44 stars 19 forks source link

BUG: VISR missing text file? #11

Closed jlmaurer closed 10 months ago

jlmaurer commented 3 years ago

Description

-When running VISR method, get an error message saying missing text file, but the file exists in the current directory.

kmaterna commented 3 years ago

@jlmaurer interesting, that doesn't happen to me, but I think I've seen it before. A few thoughts: The Python wrapper for visr creates these files in the working directory (where driver or config.txt lives):

strain_config_file = 'visr_strain.drv';
strain_data_file = 'strain_input.txt';  # can only be 20 characters long bc fortran!
strain_output_file = 'strain_output.txt';  # can only be 20 characters long bc fortran!

And then it automatically moves them into the visr output directory when it's done. When I run it, I can see them written in the same directory as the config and then I watch them get moved. Which directory do you find them in? Perhaps it's writing them in the wrong place or automatically moving them too early?

jlmaurer commented 3 years ago

Hey @kmaterna yes I think the files were left in the main execute directory for me instead of getting moved. Not sure why yet, I'll investigate and update this issue.

kmaterna commented 3 years ago

@jlmaurer when you say that, I immediately suspect differences in the behavior of subprocess.call(shell=False) or subprocess.call(shell=True) on your system vs mine. shell=False is more secure so I try to use it whenever possible, but a small subset of commands don't work unless shell=True. This is a rabbit hole I haven't wanted to enter. I'd start there with your investigation.