Config streamlining

[rdPatmore]

At present, each directory has a collection of config files: config.py, config.sh, lotus_config.sh, spice_config.sh. There is repitition between files and having bash + python handelling is convoluted.

I am proposing to streamline this process by creating a system of one centralised config file, using python. Functions within that python script can replicate the unique path definitions required by each analysis script. The process for this remapping is as follows:

• Deploy NEMO_validation as a package to simplify imports across the directories.
• Create a centralised location for the existing config files
• Move shell scripting into config.py to remove the need for manually exporting paths
  • create master function for common code across all _config.sh files
  • create a function associate with unique elements of each _config.sh file
  • create calls to master and unique functions withing analysis scripts
  • add machine detection