Application and eventually versioning problems

[roestanik]

Background

Generally, I greatly appreciate the project, many thanks for sharing. Having an untouched data set sitting in the shelf, I wanted to try to create a 3D cube out of parasound data, but I am facing several issues. At this point, the CLI commands do not run unless modified. Let me try to give some examples.

Issues

Installation

If I am not mistaken the dependencies in conda have to be installed with "conda env create -f environment.yml" instead of "conda install -f {environment.yml}".
In a Windows environment, I failed to install. In the WSL, the installation was easy.

Reading of the SGYs

This issue could eventually be version related. But the segy-sak parameters for the specification of the CMP coordinate byte location seem to be "cdp_x" and "cdp_y" (not cdpx, cdpy). Specifying "silent=True" somehow terminates the program.
Also, the reading of the segy-files sometimes failes due to header inconsistencies. I guess, this has to be solved from the data output side before.

Binning

It seems, that no lines can be generated. There is a error in line 1636 "line_id, line_id_uniq = df_nav["line"].factorize{}, TypeError: 'NoneType' is not subscriptable.

Outlook

I am thinking about moving the data processing to scripts calling your defined functions rather than using them in command line wrappers. This procedure would also allow for the generation of QC plots and further, adapted data manipulation. Or do you have other suggestions for troubleshooting?

[fwrnke]

Thanks for trying the workflow and your response!

Regarding your issues:

If I am not mistaken the dependencies in conda have to be installed with "conda env create -f environment.yml" instead of "conda install -f {environment.yml}".
In a Windows environment, I failed to install. In the WSL, the installation was easy.

That's correct, {environment.yml} is meant as a placeholder for your environment file, which can be named differently.
I developed the code on a Windows 10 machine, so I'm curious why it didn't work for you?

This issue could eventually be version related. But the segy-sak parameters for the specification of the CMP coordinate byte location seem to be "cdp_x" and "cdp_y" (not cdpx, cdpy). Specifying "silent=True" somehow terminates the program.
Also, the reading of the segy-files sometimes failes due to header inconsistencies. I guess, this has to be solved from the data output side before.

Yes, this seems like a version issue. I will have a look, thanks for the hint!

It seems, that no lines can be generated. There is a error in line 1636 "line_id, line_id_uniq = df_nav["line"].factorize{}, TypeError: 'NoneType' is not subscriptable.

Unfortunately, I cannot say much why the binning is not working for you. It seems like there is no line name present. Did you run the script using --coords_origin "aux" or "header"?

I am afraid I have only limited time at the moment but will look into these issues as soon as I can!

[roestanik]

Hey,

so I gave it an other try.

The installation also worked now on a Windows 11 machine, but the first command really needs to be changed to conda env create -f {environment.yml}, in which {environment.yml} has to be replaced with the parameter file of choice.

It seems, that no lines can be generated. There is a error in line 1636 "line_id, line_id_uniq = df_nav["line"].factorize{}, TypeError: 'NoneType' is not subscriptable.

Unfortunately, I cannot say much why the binning is not working for you. It seems like there is no line name present. Did you run the script using --coords_origin "aux" or "header"?
Working on this issue, I realized, that I set some trace header words incorrectly. It is really nicely coded, that the scripts reads all scalers and formats bytes. I just forgot to set them properly. For the documentation it would be worthwhile to note which header words are being read or used as default, so one can prepare the data better.

The first proper issue I encountered then was a naming inconsistency, one needs to rename utils_IO.py in the functions to utils_io.py, otherwise any reading of e.g. SEGYs will not work. As an alternative, I guess, one can also go through the scripts and change the import commands at the beginning of the driver scripts.

What is stopping me right now, are some static (hard coded) parameters or likewise optional parameters with default values, which can are not set while calling the main function. For example, SEGY files are being split automatically with the suffix UTM60S. Reading files without this suffix results in errors, because some lists are empty.

At this point I am not sure yet how to proceed. I have been able to do 3D binning in GLOBE Claritas, so I might just implement especially the mistie correction there. If I decided to work on your scripts and functions to make them more dynamic, I would push a modified branch to you.

Best regards and good luck!