Add flexibility for model component names (#6)

* Make the rmse writer function more general
The CIME SystemTest pgn.py will write the `init_model` variable so that
the initial condition file template can be adjusted as needed. This is
used in `rmse_writer` to send output to comp_cld.nc

* Add model component to arguments
This will add the command line / config argument for the model
component so that the input file template can be adjusted from the
json config or command line args, so this is adaptable to E3SM and CESM

* Bump version number

* Pin LIVVkit version number

Author: mkstratos

evv4esm/__init__.py

@@ -29,7 +29,7 @@
 
 from __future__ import absolute_import, division, print_function, unicode_literals
 
-__version_info__ = (0, 2, 3)
+__version_info__ = (0, 2, 4)
 __version__ = '.'.join(str(vi) for vi in __version_info__)
 
 PASS_COLOR = '#389933'

evv4esm/extensions/ks.py

@@ -123,6 +123,10 @@ def parse_args(args=None):
                         default=os.getcwd(),
                         help='Image output location.')
 
+    parser.add_argument('--component',
+                        default='eam',
+                        help='Model component name (e.g. eam, cam, ...)')
+
     args, _ = parser.parse_known_args(args)
 
     # use config file arguments, but override with command line arguments
@@ -199,8 +203,8 @@ def case_files(args):
         key1 += '1'
         key2 += '2'
 
-    f_sets = {key1: e3sm.component_monthly_files(args.test_dir, 'eam', args.ninst),
-              key2: e3sm.component_monthly_files(args.ref_dir, 'eam', args.ninst)}
+    f_sets = {key1: e3sm.component_monthly_files(args.test_dir, args.component, args.ninst),
+              key2: e3sm.component_monthly_files(args.ref_dir, args.component, args.ninst)}
 
     for key in f_sets:
         # Require case files for at least the last 12 months.

evv4esm/extensions/pg.py

@@ -109,7 +109,7 @@ def _sub2instance(initial_condition, perturbation_index, total_perturbations):
     return instance
 
 
-def rmse_writer(file_name, rmse, perturbation_names, perturbation_variables, init_file_template):
+def rmse_writer(file_name, rmse, perturbation_names, perturbation_variables, init_file_template, model_name):
     """
     Opens and writes a netcdf file for PGE curves
     This function is here purely to avoid duplicate
@@ -139,7 +139,7 @@ def rmse_writer(file_name, rmse, perturbation_names, perturbation_variables, ini
 
         for icond in range(0, ninit):
             # NOTE: Zero vs One based indexing
-            nc_init_cond[icond] = init_file_template.format('eam', 'i', icond+1)
+            nc_init_cond[icond] = init_file_template.format(model_name, 'i', icond+1)
 
 
 def variables_rmse(ifile_test, ifile_cntl, var_list, var_pefix=''):
@@ -214,13 +214,13 @@ def main(args):
                 continue
             iinst_ctrl = _sub2instance(icond, 0, nprt)
             ifile_ctrl = os.path.join(args.ref_dir,
-                                      args.instance_file_template.format('', iinst_ctrl, '_woprt'))
+                                      args.instance_file_template.format('', args.component, iinst_ctrl, '_woprt'))
             # logger.debug("PGN_INFO:CNTL_TST:" + ifile_cntl)
 
             iinst_test = _sub2instance(icond, iprt, nprt)
             ifile_test = os.path.join(args.test_dir,
                                       args.instance_file_template.format(
-                                              args.test_case + '.', iinst_test, '_' + prt_name))
+                                              args.test_case + '.', args.component, iinst_test, '_' + prt_name))
             # logger.debug("PGN_INFO:TEST_TST:" + ifile_test)
 
             prt_rmse[prt_name] = variables_rmse(ifile_test, ifile_ctrl, args.variables, 't_')
@@ -231,7 +231,7 @@ def main(args):
     comp_rmse = np.reshape(rmse.RMSE.values, (args.ninit, nprt-1, nvar))
 
     rmse_writer(os.path.join(args.test_dir, 'comp_cld.nc'),
-                comp_rmse, args.perturbations.keys(), args.variables, args.init_file_template)
+                comp_rmse, args.perturbations.keys(), args.variables, args.init_file_template, args.init_model)
 
     details = OrderedDict()
     with Dataset(os.path.join(args.ref_dir, args.pge_cld)) as ref_cld:

evv4esm/extensions/tsc.py

@@ -163,18 +163,18 @@ def main(args):
         args.test_case += '1'
         args.ref_case += '2'
 
-    file_search_glob = '{d}/*eam_????.h0.0001-01-01-?????.nc.{s}'
+    file_search_glob = '{d}/*{c}_????.h0.0001-01-01-?????.nc.{s}'
     truth_ens = {instance: list(files) for instance, files in groupby(
-            sorted(glob.glob(file_search_glob.format(d=args.ref_dir, s='DT0001'))),
-            key=lambda f: e3sm.component_file_instance('eam', f))}
+            sorted(glob.glob(file_search_glob.format(d=args.ref_dir, c=args.component, s='DT0001'))),
+            key=lambda f: e3sm.component_file_instance(args.baseline_component, f))}
 
     ref_ens = {instance: list(files) for instance, files in groupby(
-            sorted(glob.glob(file_search_glob.format(d=args.ref_dir, s='DT0002'))),
-            key=lambda f: e3sm.component_file_instance('eam', f))}
+            sorted(glob.glob(file_search_glob.format(d=args.ref_dir, c=args.component, s='DT0002'))),
+            key=lambda f: e3sm.component_file_instance(args.baseline_component, f))}
 
     test_ens = {instance: list(files) for instance, files in groupby(
-            sorted(glob.glob(file_search_glob.format(d=args.test_dir, s='DT0002'))),
-            key=lambda f: e3sm.component_file_instance('eam', f))}
+            sorted(glob.glob(file_search_glob.format(d=args.test_dir, c=args.component, s='DT0002'))),
+            key=lambda f: e3sm.component_file_instance(args.component, f))}
 
     # So, we want a pandas dataframe that will have the columns :
     #     (test/ref, ensemble, seconds, l2_global, l2_land, l2_ocean)

setup.py

@@ -79,7 +79,7 @@
           'numpy',
           'scipy',
           'pandas',
-          'livvkit',
+          'livvkit==2.1.6',
           'netCDF4',
           'matplotlib',
           'pybtex',