UnboundLocalError: local variable 'log' referenced before assignment about gwidecodeml HOT 4 CLOSED

Dear Niraj,

Thank you for using GwideCodeML.

It was indeed a bug in the code that was already fixed. You can clone again the repository and type python setup.py install to get the package with the newest changes.

However, this bug appeared when the function is not able to get a LnL value of the tested hypothesis after running codeml. This means that codeml did not successfully run with any of your provided alignment files. In the fixed version, a warning is printed in the log file (gwidecodeml.log) when this happens. You can run again your analysis and see which are the problematic files. If you don't know why codeml is not running with your files you can send me an example of the alignment files used together with a tree file and branch file and I will tell you which is exactly the problem.

Best,
Laura

from gwidecodeml.

Hi Laura,

I think you commented out the code that runs codeml in parallel.

#pool = mp.Pool(t) # number of# threads
#pool.map(utils.run_codeml, [ctl for ctl in gene_names])
#pool.close()
#os.chdir(working_dir)

Can you please check? I am getting the following message:
2021-02-11 14:58:20,358 :: INFO :: Starting analysis of 3269 alignments
2021-02-11 14:58:46,294 :: INFO :: Control files successfully created, GWidecodeml is ready for codeml performance
2021-02-11 14:58:46,294 :: INFO :: Codeml has finished, output files successfully created. Running LRTs...
2021-02-11 14:58:46,294 :: INFO :: DEBUGGING

But I DO NOT see any codeml output. It is possibly because the code for running codeml has been commented out.

UPDATE:
I removed the comment and set it up again. It worked!

Best,
Niraj

from gwidecodeml.

Dear Niraj,

I want to apologize because I've totally mixed up the versions when I was trying to fix the bug. The code has been updated.

Regarding the error of the branches file, please, make sure that your text file does not contain any blank line.

Thank you for reporting and again, sorry for the inconvenience caused.

Laura

from gwidecodeml.

Hi Laura,

No problem! Thank you for your good work!

Best,
Niraj

from gwidecodeml.