Implement dynamic setup of default datasets directory for executing gp_baseline.py

[digitallance]

gp_baseline.py script assumes there is a datasets directory just 1 level up from cwd.

Code line: shutil.copy('../datasets/meshcs_to_gocc.csv', os.getcwd()+'/datasets') uses the os.getcwd() directory path only in one instance (destination). One might as well use './datasets' since the code already switched the working directory to the command line provided path. The destination path does exist since it was created just earlier but the source path is not dynamic and may not exist.

One may need a second command line argument with the location of the source datasets directory. Alternatively, the script has to be launched from a specific directory with a datasets directory available and proper directory changes have to be done by the script.

Another issue is that if you start the script from the wrong location, equiv module fails to load since it assumed a relative path to resources as well.

In principal, the config file could have locations of assumed resources provided.

[digitallance]

Modified gp_baseline.py to allow execution from any location with proper path checking for loading and copying files from datasets directory by this script and by equiv.py module on import.