No output from compute_trace.py

[EEPuckett]

Did anyone need to edit the compute_trace.py? It's unclear to me how it should work.
First- there does not appear to be any print function for output_filename

Second- The input for -prefix is very confusing to me. Help says it is an integer for Ne. However, in the code it looks like the output parameter from convert_to_tskit. If within convert_trace the following is defined:
file_name = f"{prefix}_{index}.trees"
Then I don't understand how prefix is an integer if I did not use an integer as the name for my .trees files.

With thanks, Emily

[moshejasper]

Hi @EEPuckett

(writing as a fellow user)

Based on my inspection of the python code, this script has not been completed & is not yet ready for use from the command line.

There is no correlation between the -prefix listed in the help file and the {prefix} you found in code.
Assuming you are wanting to compute 'diversity' traces, It would be possible to just make a python script that imports this file into python, then run the 'diversity_fit_trace' function... but more work is needed here before this is operational.

[SillySabertooth]

Hello there,

(writing as another fellow user)

I'm not experienced in python, but it seems that compute_trace.py is missing only a few final pieces. I allowed myself to try to add them, and it seems it's working now. I haven't added anything into the working code (only changed the default append function into numpy append), but i) I changed the misnamed argument; ii) added calling of the functions and saving the results. I run the adjusted script on my singer output, and it seems the code works (and the output seems to show that I need to add more iterations/samples to my singer workflow)

Running it as provided in the manual:
python singer-0.1.8/compute_trace.adj.py -prefix chr$chrom.tskit_files -m 1.25e-8 -start_index 0 -end_index 100 -output_filename chr$chrom.traces

All the best

compute_trace.adj.py