Is your feature request related to a problem? Please describe.
It can be very hard for me to locate generated files in the experiments folder, especially when many files are generated in the same folder.
Describe the solution you'd like
Give the user the option to change the instance folder name (current it's always the timestamp of when the SimOpt instance opened)
Organize all files into subfolders, such as Instance_Folder/Experiment_Name/Problem_Name/Solver_Name.pickle. This structure makes it easy to identify what solvers go to what problems and groups all the data needed for post-normalizing a problem in a single folder.
Look into removing the Ruby script input/output files. In theory, all this data could be handled by the subprocess call without needing to clutter the directory (but could still be saved to the directory if desired). Alternatively, the Ruby input/output file(s) could be set to automatically delete themselves if not desired.
Add an I/O manager singleton to handle all file I/O ops. This would be a nice QOL improvement as file management would be abstracted away from simulation code.
Is your feature request related to a problem? Please describe. It can be very hard for me to locate generated files in the experiments folder, especially when many files are generated in the same folder.
Describe the solution you'd like
Instance_Folder/Experiment_Name/Problem_Name/Solver_Name.pickle
. This structure makes it easy to identify what solvers go to what problems and groups all the data needed for post-normalizing a problem in a single folder.