Added disc_on kwarg through cli to all_sky

Author: bnb32

nsrdb/all_sky/all_sky.py

@@ -49,7 +49,7 @@
 def all_sky(alpha, aod, asymmetry, cloud_type, cld_opd_dcomp, cld_reff_dcomp,
             ozone, solar_zenith_angle, ssa, surface_albedo, surface_pressure,
             time_index, total_precipitable_water, cloud_fill_flag=None,
-            variability_kwargs=None, scale_outputs=True, farmsdni=True):
+            variability_kwargs=None, scale_outputs=True, disc_on=False):
     """Calculate the all-sky irradiance.
     Updated by Yu Xie on 3/29/2023 to compute DNI by FARMS-DNI.
 
@@ -109,8 +109,8 @@ def all_sky(alpha, aod, asymmetry, cloud_type, cld_opd_dcomp, cld_reff_dcomp,
         nsrdb.all_sky.utilities.cloud_variability for kwarg definitions.
     scale_outputs : bool
         Flag to safely scale and dtype convert output arrays.
-    farmsdni : bool
-        Compute cloudy-sky DNI using FARMS-DNI (True) or DISC (False).
+    disc_on : bool
+        Compute cloudy-sky DNI using FARMS-DNI (False) or DISC (True).
 
     Returns
     -------
@@ -163,16 +163,15 @@ def all_sky(alpha, aod, asymmetry, cloud_type, cld_opd_dcomp, cld_reff_dcomp,
 
     # use FARMS to calculate cloudy GHI
     ghi, dni_farmsdni, _ = farms(tau=cld_opd_dcomp,
-                                    cloud_type=cloud_type,
-                                    cloud_effective_radius=cld_reff_dcomp,
-                                    solar_zenith_angle=solar_zenith_angle,
-                                    radius=radius,
-                                    Tuuclr=Tuuclr,
-                                    Ruuclr=rest_data.Ruuclr,
-                                    Tddclr=rest_data.Tddclr,
-                                    Tduclr=rest_data.Tduclr,
-                                    albedo=surface_albedo,
-                                    )
+                                 cloud_type=cloud_type,
+                                 cloud_effective_radius=cld_reff_dcomp,
+                                 solar_zenith_angle=solar_zenith_angle,
+                                 radius=radius,
+                                 Tuuclr=Tuuclr,
+                                 Ruuclr=rest_data.Ruuclr,
+                                 Tddclr=rest_data.Tddclr,
+                                 Tduclr=rest_data.Tduclr,
+                                 albedo=surface_albedo)
 
     # merge the clearsky and cloudy irradiance into all-sky irradiance
     ghi = merge_rest_farms(rest_data.ghi, ghi, cloud_type)
@@ -182,16 +181,13 @@ def all_sky(alpha, aod, asymmetry, cloud_type, cld_opd_dcomp, cld_reff_dcomp,
         ghi = cloud_variability(ghi, rest_data.ghi, cloud_type,
                                 **variability_kwargs)
 
-    # calculate the DNI using the DISC model
-    dni = disc(ghi, solar_zenith_angle, doy=time_index.dayofyear.values,
-               pressure=surface_pressure)
-
     # merge the clearsky and cloudy irradiance into all-sky irradiance
-    #
-    if farmsdni:
-        dni = merge_rest_farms(rest_data.dni, dni_farmsdni, cloud_type)
-    else:
+    if disc_on:
+        dni = disc(ghi, solar_zenith_angle, doy=time_index.dayofyear.values,
+                   pressure=surface_pressure)
         dni = merge_rest_farms(rest_data.dni, dni, cloud_type)
+    else:
+        dni = merge_rest_farms(rest_data.dni, dni_farmsdni, cloud_type)
 
     # make a fill flag where bad data exists in the GHI irradiance
     fill_flag = make_fill_flag(ghi, rest_data.ghi, cloud_type, missing_props,
@@ -239,7 +235,7 @@ def all_sky(alpha, aod, asymmetry, cloud_type, cld_opd_dcomp, cld_reff_dcomp,
     return output
 
 
-def all_sky_h5(f_source, rows=slice(None), cols=slice(None)):
+def all_sky_h5(f_source, rows=slice(None), cols=slice(None), disc_on=False):
     """Run all-sky from .h5 files.
 
     Parameters
@@ -251,6 +247,8 @@ def all_sky_h5(f_source, rows=slice(None), cols=slice(None)):
         Subset of rows to run.
     cols : slice
         Subset of columns to run.
+    disc_on : bool
+        Compute cloudy sky dni with disc model or farms-dni model.
 
     Returns
     -------
@@ -275,6 +273,7 @@ def all_sky_h5(f_source, rows=slice(None), cols=slice(None)):
         all_sky_input = {dset: source[dset, rows, cols]
                          for dset in ALL_SKY_ARGS}
         all_sky_input['time_index'] = source.time_index[rows].tz_convert(None)
+        all_sky_input['disc_on'] = disc_on
 
     try:
         out = all_sky(**all_sky_input)
@@ -286,7 +285,7 @@ def all_sky_h5(f_source, rows=slice(None), cols=slice(None)):
 
 
 def all_sky_h5_parallel(f_source, rows=slice(None), cols=slice(None),
-                        col_chunk=10, max_workers=None):
+                        col_chunk=10, max_workers=None, disc_on=False):
     """Run all-sky from .h5 files.
 
     Parameters
@@ -304,6 +303,8 @@ def all_sky_h5_parallel(f_source, rows=slice(None), cols=slice(None),
         significantly faster.
     max_workers : int | None
         Number of workers to run in parallel.
+    disc_on : bool
+        Compute cloudy sky dni with disc model or farms-dni model.
 
     Returns
     -------
@@ -361,7 +362,7 @@ def all_sky_h5_parallel(f_source, rows=slice(None), cols=slice(None),
 
     loggers = ['farms', 'nsrdb', 'rest2', 'rex']
     with SpawnProcessPool(max_workers=max_workers, loggers=loggers) as exe:
-        futures = {exe.submit(all_sky_h5, f_source, rows=rows,
+        futures = {exe.submit(all_sky_h5, f_source, rows=rows, disc_on=disc_on,
                               cols=c_slices_all[c]): c for c in c_range}
 
         for future in as_completed(futures):

nsrdb/cli.py

@@ -930,8 +930,11 @@ def ml_cloud_fill(ctx, date, fill_all, model_path, col_chunk, max_workers):
               help='Chunking method to run all sky one column chunk at a time '
               'to reduce memory requirements. This is an integer specifying '
               'how many columns to work on at one time.')
+@click.option('--disc_on', '-do', type=bool, required=False, default=False,
+              help='Whether to run compute cloudy sky dni with the disc model '
+              '(True) or the farms-dni model (False).')
 @click.pass_context
-def all_sky(ctx, i_chunk, col_chunk):
+def all_sky(ctx, i_chunk, col_chunk, disc_on):
     """Run allsky for a single chunked file"""
 
     name = ctx.obj['NAME']
@@ -945,9 +948,9 @@ def all_sky(ctx, i_chunk, col_chunk):
     log_file = 'all_sky/all_sky_{}.log'.format(i_chunk)
     fun_str = 'NSRDB.run_all_sky'
     arg_str = ('"{}", {}, "{}", freq="{}", i_chunk={}, col_chunk={}, '
-               'log_file="{}", log_level="{}", job_name="{}"'
+               'disc_on={}, log_file="{}", log_level="{}", job_name="{}"'
                .format(out_dir, year, nsrdb_grid, nsrdb_freq, i_chunk,
-                       col_chunk, log_file, log_level, name))
+                       col_chunk, disc_on, log_file, log_level, name))
     if var_meta is not None:
         arg_str += ', var_meta="{}"'.format(var_meta)
     ctx.obj['IMPORT_STR'] = 'from nsrdb.nsrdb import NSRDB'
@@ -964,8 +967,11 @@ def all_sky(ctx, i_chunk, col_chunk):
               help='Chunking method to run all sky one column chunk at a time '
               'to reduce memory requirements. This is an integer specifying '
               'how many columns to work on at one time.')
+@click.option('--disc_on', '-do', type=bool, required=False, default=False,
+              help='Whether to run compute cloudy sky dni with the disc model '
+              '(True) or the farms-dni model (False).')
 @click.pass_context
-def daily_all_sky(ctx, date, col_chunk):
+def daily_all_sky(ctx, date, col_chunk, disc_on):
     """Run allsky for a single day using daily data model output files as
     source data"""
 
@@ -979,10 +985,10 @@ def daily_all_sky(ctx, date, col_chunk):
 
     log_file = 'all_sky/all_sky_{}.log'.format(date)
     fun_str = 'NSRDB.run_daily_all_sky'
-    arg_str = ('"{}", {}, "{}", "{}", freq="{}", col_chunk={}, '
+    arg_str = ('"{}", {}, "{}", "{}", freq="{}", col_chunk={}, disc_on={}, '
                'log_file="{}", log_level="{}", job_name="{}"'
                .format(out_dir, year, nsrdb_grid, date, nsrdb_freq, col_chunk,
-                       log_file, log_level, name))
+                       disc_on, log_file, log_level, name))
     if var_meta is not None:
         arg_str += ', var_meta="{}"'.format(var_meta)
     ctx.obj['IMPORT_STR'] = 'from nsrdb.nsrdb import NSRDB'

nsrdb/nsrdb.py

@@ -1405,7 +1405,8 @@ def ml_cloud_fill(cls, out_dir, date, fill_all=False,
     def run_all_sky(cls, out_dir, year, grid, freq='5min', var_meta=None,
                     col_chunk=10, rows=slice(None), cols=slice(None),
                     max_workers=None, log_level='DEBUG',
-                    log_file='all_sky.log', i_chunk=None, job_name=None):
+                    log_file='all_sky.log', i_chunk=None, job_name=None,
+                    disc_on=False):
         """Run the all-sky physics model from collected .h5 files
 
         Parameters
@@ -1442,6 +1443,9 @@ def run_all_sky(cls, out_dir, year, grid, freq='5min', var_meta=None,
             Enumerated file index if running on site chunk.
         job_name : str
             Optional name for pipeline and status identification.
+        disc_on : bool
+            Compute cloudy sky dni with the disc model (True) or the farms-dni
+            model (False)
         """
         t0 = time.time()
         nsrdb = cls(out_dir, year, grid, freq=freq, var_meta=var_meta)
@@ -1476,10 +1480,10 @@ def run_all_sky(cls, out_dir, year, grid, freq='5min', var_meta=None,
         if max_workers != 1:
             out = all_sky_h5_parallel(f_source, rows=rows, cols=cols,
                                       max_workers=max_workers,
-                                      col_chunk=col_chunk)
+                                      col_chunk=col_chunk, disc_on=disc_on)
         else:
             out = all_sky_h5(f_source, rows=rows, cols=cols,
-                             col_chunk=col_chunk)
+                             col_chunk=col_chunk, disc_on=disc_on)
 
         logger.info('Finished all-sky. Writing to: {}'.format(f_out))
         with Outputs(f_out, mode='a') as f:
@@ -1511,7 +1515,8 @@ def run_daily_all_sky(cls, out_dir, year, grid, date, freq='5min',
                           var_meta=None, col_chunk=500,
                           rows=slice(None), cols=slice(None),
                           max_workers=None, log_level='DEBUG',
-                          log_file='all_sky.log', job_name=None):
+                          log_file='all_sky.log', job_name=None,
+                          disc_on=False):
         """Run the all-sky physics model from daily data model output files.
 
         Parameters
@@ -1549,6 +1554,9 @@ def run_daily_all_sky(cls, out_dir, year, grid, date, freq='5min',
             File to log to. Will be put in output directory.
         job_name : str
             Optional name for pipeline and status identification.
+        disc_on : bool
+            Compute cloudy sky dni with the disc model (True) or the farms-dni
+            model (False)
         """
         t0 = time.time()
         assert len(str(date)) == 8
@@ -1564,9 +1572,9 @@ def run_daily_all_sky(cls, out_dir, year, grid, date, freq='5min',
         if max_workers != 1:
             out = all_sky_h5_parallel(f_source, rows=rows, cols=cols,
                                       max_workers=max_workers,
-                                      col_chunk=col_chunk)
+                                      col_chunk=col_chunk, disc_on=disc_on)
         else:
-            out = all_sky_h5(f_source, rows=rows, cols=cols)
+            out = all_sky_h5(f_source, rows=rows, cols=cols, disc_on=disc_on)
 
         logger.info('Finished all-sky compute.')
         for dset, arr in out.items():
@@ -1597,7 +1605,7 @@ def run_daily_all_sky(cls, out_dir, year, grid, date, freq='5min',
     def run_full(cls, date, grid, freq, var_meta=None, factory_kwargs=None,
                  fill_all=False, model_path=None, dist_lim=1.0,
                  max_workers=None, low_mem=False,
-                 log_file=None, log_level='INFO'):
+                 log_file=None, log_level='INFO', disc_on=False):
         """Run the full nsrdb pipeline in-memory using serial compute.
 
         Parameters
@@ -1648,6 +1656,9 @@ def run_full(cls, date, grid, freq, var_meta=None, factory_kwargs=None,
             initialized.
         log_file : str
             File to log to. Will be put in output directory.
+        disc_on : bool
+            Compute cloudy sky dni with the disc model (True) or the farms-dni
+            model (False)
 
         Returns
         -------
@@ -1693,6 +1704,7 @@ def run_full(cls, date, grid, freq, var_meta=None, factory_kwargs=None,
                           if k in ALL_SKY_ARGS}
         all_sky_inputs['time_index'] = data_model.nsrdb_ti
         all_sky_inputs['scale_outputs'] = False
+        all_sky_inputs['disc_on'] = disc_on
         logger.info('Running NSRDB All-Sky.')
         all_sky_out = all_sky(**all_sky_inputs)
         for k, v in all_sky_out.items():