Feature request: Add benchmark results paths and other related info to the the config object

[yantosca]

Name and Institution (Required)

Name: Bob Yantosca
Institution: Harvard + GCST

Confirm you have reviewed the following documentation

• Contributing guidelines

New GCPy feature or discussion

In each of the scripts run_benchmark.py, run_1yr_fullchem_benchmark.py, run_1yr_tt_benchmark.py, there is repeated code similar to this that defines the various data directories and other quantities:

gcpy/gcpy/benchmark/modules/run_1yr_fullchem_benchmark.py

Lines 100 to 271 in eb0055b

	# ======================================================================
	# Data directories
	# ======================================================================
	# Diagnostics file directory paths
	gcc_vs_gcc_refdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["ref"]["gcc"]["dir"],
	config["data"]["ref"]["gcc"]["outputs_subdir"],
	)
	gcc_vs_gcc_devdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gcc"]["dir"],
	config["data"]["dev"]["gcc"]["outputs_subdir"],
	)
	gchp_vs_gcc_refdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gcc"]["dir"],
	config["data"]["dev"]["gcc"]["outputs_subdir"],
	)
	gchp_vs_gcc_devdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gchp"]["dir"],
	config["data"]["dev"]["gchp"]["outputs_subdir"],
	)
	gchp_vs_gchp_refdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["ref"]["gchp"]["dir"],
	config["data"]["ref"]["gchp"]["outputs_subdir"],
	)
	gchp_vs_gchp_devdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gchp"]["dir"],
	config["data"]["dev"]["gchp"]["outputs_subdir"],
	)
	# Restart file directory paths
	gcc_vs_gcc_refrstdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["ref"]["gcc"]["dir"],
	config["data"]["ref"]["gcc"]["restarts_subdir"]
	)
	gcc_vs_gcc_devrstdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gcc"]["dir"],
	config["data"]["dev"]["gcc"]["restarts_subdir"]
	)
	gchp_vs_gcc_refrstdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gcc"]["dir"],
	config["data"]["dev"]["gcc"]["restarts_subdir"]
	)
	gchp_vs_gcc_devrstdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gchp"]["dir"],
	config["data"]["dev"]["gchp"]["restarts_subdir"]
	)
	gchp_vs_gchp_refrstdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["ref"]["gchp"]["dir"],
	config["data"]["ref"]["gchp"]["restarts_subdir"]
	)
	gchp_vs_gchp_devrstdir = os.path.join(
	config["paths"]["main_dir"],
	config["data"]["dev"]["gchp"]["dir"],
	config["data"]["dev"]["gchp"]["restarts_subdir"]
	)
	# Directories where plots & tables will be created
	mainresultsdir = os.path.join(
	config["paths"]["results_dir"]
	)
	gcc_vs_gcc_resultsdir = os.path.join(
	mainresultsdir,
	config["options"]["comparisons"]["gcc_vs_gcc"]["dir"]
	)
	gchp_vs_gcc_resultsdir = os.path.join(
	mainresultsdir,
	config["options"]["comparisons"]["gchp_vs_gcc"]["dir"]
	)
	gchp_vs_gchp_resultsdir = os.path.join(
	mainresultsdir,
	config["options"]["comparisons"]["gchp_vs_gchp"]["dir"]
	)
	diff_of_diffs_resultsdir = os.path.join(
	mainresultsdir,
	"GCHP_GCC_diff_of_diffs"
	)
	# Make copy of benchmark script in results directory
	if not os.path.exists(mainresultsdir):
	os.mkdir(mainresultsdir)
	curfile = os.path.realpath(__file__)
	dest = os.path.join(mainresultsdir, curfile.split("/")[-1])
	if not os.path.exists(dest):
	copyfile(curfile, dest)
	# Create results directories that don't exist,
	# and place a copy of this file in each results directory
	results_list = [
	gcc_vs_gcc_resultsdir,
	gchp_vs_gchp_resultsdir,
	gchp_vs_gcc_resultsdir,
	diff_of_diffs_resultsdir
	]
	comparisons_list = [
	config["options"]["comparisons"]["gcc_vs_gcc"]["run"],
	config["options"]["comparisons"]["gchp_vs_gchp"]["run"],
	config["options"]["comparisons"]["gchp_vs_gcc"]["run"],
	config["options"]["comparisons"]["gchp_vs_gcc_diff_of_diffs"]["run"]
	]
	for resdir, plotting_type in zip(results_list, comparisons_list):
	if plotting_type and not os.path.exists(resdir):
	os.mkdir(resdir)
	if resdir in results_list:
	curfile = os.path.realpath(__file__)
	dest = os.path.join(resdir, curfile.split("/")[-1])
	if not os.path.exists(dest):
	copyfile(curfile, dest)
	# Tables directories
	gcc_vs_gcc_tablesdir = os.path.join(
	gcc_vs_gcc_resultsdir,
	config["options"]["comparisons"]["gcc_vs_gcc"]["tables_subdir"],
	)
	gchp_vs_gcc_tablesdir = os.path.join(
	gchp_vs_gcc_resultsdir,
	config["options"]["comparisons"]["gchp_vs_gcc"]["tables_subdir"],
	)
	gchp_vs_gchp_tablesdir = os.path.join(
	gchp_vs_gchp_resultsdir,
	config["options"]["comparisons"]["gchp_vs_gchp"]["tables_subdir"],
	)
	## Budget directories
	#gcc_vs_gcc_budgetdir = os.path.join(gcc_vs_gcc_resultsdir, "Budget")
	#gchp_vs_gcc_budgetdir = os.path.join(gchp_vs_gcc_resultsdir, "Budget")
	#gchp_vs_gchp_budgetdir = os.path.join(gchp_vs_gchp_resultsdir, "Budget")
	# Models vs. observations directories
	gcc_vs_gcc_models_vs_obs_dir = os.path.join(
	gcc_vs_gcc_resultsdir, "ModelVsObs"
	)
	gchp_vs_gcc_models_vs_obs_dir = os.path.join(
	gchp_vs_gcc_resultsdir, "ModelVsObs"
	)
	gchp_vs_gchp_models_vs_obs_dir = os.path.join(
	gchp_vs_gchp_resultsdir, "ModelVsObs"
	)
	# ======================================================================
	# Plot title strings
	# ======================================================================
	gcc_vs_gcc_refstr = config["data"]["ref"]["gcc"]["version"]
	gcc_vs_gcc_devstr = config["data"]["dev"]["gcc"]["version"]
	gchp_vs_gcc_refstr = config["data"]["dev"]["gcc"]["version"]
	gchp_vs_gcc_devstr = config["data"]["dev"]["gchp"]["version"]
	gchp_vs_gchp_refstr = config["data"]["ref"]["gchp"]["version"]
	gchp_vs_gchp_devstr = config["data"]["dev"]["gchp"]["version"]
	diff_of_diffs_refstr = bmk.diff_of_diffs_toprow_title(config, "gcc")
	diff_of_diffs_devstr = bmk.diff_of_diffs_toprow_title(config, "gchp")
	# ======================================================================
	# Observational data files
	# ======================================================================
	sondes_data_file = os.path.join(
	config["paths"]["obs_data"]["sondes"]["data_dir"],
	config["paths"]["obs_data"]["sondes"]["data_file"],
	)
	sondes_site_file = os.path.join(
	config["paths"]["obs_data"]["sondes"]["data_dir"],
	config["paths"]["obs_data"]["sondes"]["site_file"],
	)

These could be folded into the config object when the YAML configuration file is read. This would clean up the run*benchmark..py scripts quite a bit.

Also see this issue:

• [FEATURE REQUEST]: Further DRY out various benchmark plotting scripts #141

NOTE: "DRY" = "Don't repeat yourself"

Tagging @lizziel @msulprizio