Feature Request: logrx working with parallel processing

[bms63]

Feature Idea

Investigate how logrx can work with parallel processing

Relevant Input

No response

Relevant Output

No response

Reproducible Example/Pseudo Code

No response

[kodesiba]

I do a bunch of parallel processing and can look into this. We may want to outline what use case(s) we want to check first since I can think of a few.

[danielmcd]

The case where I ran into the limitation was in the context of running many parallel runs of the same program. Imagine a simulation where you want to run it many times with different random seeds, so you can compute statistics across the different scenarios at the end. The setup passed the run number to the program as a command line argument. The issue is that logrx is assuming that you want to execute a file and does not have a way to pass arguments to the file.

I think the issue can be generalized to allowing execution of R scripts that accept command arguments via the commandArgs() function and add an optional parameter to logrx::axecute() that allows you to pass a list of arguments.

[AbhaySThakur]

One use case we have is to run large number of R scripts with inhouse package I developed (which internally uses logrx). Now building a wrapper to make the process conditionally parallel for some files. I am planning to close this in next few weeks, if logrx can release this in next cycle then will simplify the code greatly at my end.