Several fixes

README.md

@@ -210,7 +210,7 @@ Where to store grids extracted from baseline datasets. When `baseline.name` is `
 
 ### paths.grids.forecast
 
-Where to store grids extracted from netCDF or Zarr forecast datasets. GRIB forecast datasets currently must be local, and grids are processed directly from their containing files without being extracted into separate files.
+Where to store grids extracted from netCDF or Zarr forecast datasets. GRIB forecast datasets currently must be local, and grids are processed directly from their containing files without being extracted into separate files. This value is required for netCDF and Zarr datasets, and ignored for local GRIB datasets.
 
 ### paths.grids.truth
 

recipe/meta.json

@@ -48,5 +48,5 @@
       "zarr ==3.1.*"
     ]
   },
-  "version": "0.7.0"
+  "version": "0.7.1"
 }

src/wxvx/metconf.py

@@ -226,8 +226,11 @@ def _top(k: str, v: Any, level: int) -> list[str]:
         case MET.quality_mark_thresh:
             return _kvpair(k, _bare(v), level)
         # Scalar: quoted.
-        case MET.model | MET.obtype | MET.output_prefix | MET.tmp_dir:
+        case MET.obtype | MET.output_prefix | MET.tmp_dir:
             return _kvpair(k, _quoted(v), level)
+        # Scalar: quoted with no whitespace.
+        case MET.model:
+            return _kvpair(k, _quoted(v.replace(" ", "_")), level)
         # Sequence: quoted.
         case MET.message_type | MET.obs_bufr_var:
             return _sequence(k, v, _quoted, level)

src/wxvx/resources/config.jsonschema

@@ -334,9 +334,6 @@
               "type": "string"
             }
           },
-          "required": [
-            "forecast"
-          ],
           "type": "object"
         },
         "obs": {

src/wxvx/resources/info.json

@@ -1,4 +1,4 @@
 {
   "buildnum": "0",
-  "version": "0.7.0"
+  "version": "0.7.1"
 }

src/wxvx/strings.py

@@ -132,6 +132,7 @@ class _S(_ValsMatchKeys):
     Strings defined by wxvx, plus strings from various other sources.
     """
 
+    GDAS: str = _
     GFS: str = _
     HRRR: str = _
     OBS: str = _

src/wxvx/tests/conftest.py

@@ -81,7 +81,7 @@ def config_data():
                 [21.138123, 275.0],
                 [21.138123, 225.90452027],
             ],
-            S.name: "Forecast",
+            S.name: "Forecast Model",
             S.path: "/path/to/forecast-{{ yyyymmdd }}-{{ hh }}-{{ '%03d' % fh }}.nc",
             S.projection: {
                 "a": 6371229,

src/wxvx/tests/test_schema.py

@@ -271,12 +271,8 @@ def test_schema_paths_grids(config_data, fs, logged):
     for key in [S.forecast, S.truth]:
         assert not ok(with_set(config, None, key))
         assert logged("None is not of type 'string'")
-    # Some values are required:
-    for key in [S.forecast]:
-        assert not ok(with_del(config, key))
-        assert logged(f"'{key}' is a required property")
     # Some values are optional:
-    for key in [S.baseline]:
+    for key in [S.baseline, S.forecast]:
         assert ok(with_del(config, key))
 
 

src/wxvx/tests/test_types.py

@@ -248,7 +248,7 @@ def test_types_Forecast(config_data, forecast):
     assert obj.coords.time.inittime == "time"
     assert obj.coords.time.leadtime == "lead_time"
     assert obj.format is None
-    assert obj.name == "Forecast"
+    assert obj.name == "Forecast Model"
     assert obj.path == "/path/to/forecast-{{ yyyymmdd }}-{{ hh }}-{{ '%03d' % fh }}.nc"
     cfg = config_data[S.forecast]
     other1 = types.Forecast(**cfg)

src/wxvx/tests/test_util.py

@@ -75,11 +75,15 @@ def test_util_classify_data_format__pass_zarr(tmp_path):
         ("http://example.com/path/to/gfs.t00z.pgrb2.0p25.f001", util.Proximity.REMOTE),
         ("file:///path/to/gfs.t00z.pgrb2.0p25.f001", util.Proximity.LOCAL),
         ("/path/to/gfs.t00z.pgrb2.0p25.f001", util.Proximity.LOCAL),
+        ("gfs.t00z.pgrb2.0p25.f001", util.Proximity.LOCAL),
     ],
 )
 def test_workflow_classify_url(expected_scheme, url):
-    scheme, _ = util.classify_url(url)
+    scheme, new = util.classify_url(url)
     assert scheme == expected_scheme
+    if scheme == util.Proximity.LOCAL:
+        assert isinstance(new, Path)
+        assert new.is_absolute()
 
 
 def test_workflow_classify_url_unsupported():

src/wxvx/tests/test_variables.py

@@ -248,7 +248,7 @@ class C1(B): ...
     class C2(B): ...
 
     assert variables.model_names(A) == {"B", "C1", "C2"}
-    assert variables.model_names() == {S.GFS, S.HRRR, S.PREPBUFR}
+    assert variables.model_names() == {S.GDAS, S.GFS, S.HRRR, S.PREPBUFR}
 
 
 def test_variables__da_val__fail_unparesable(da_flat):

src/wxvx/tests/test_workflow.py

@@ -275,7 +275,7 @@ def test_workflow__config_point_stat__atm(c, fakefs, fmt, testvars, tidy):
         val = "ADPSFC";
       }
     ];
-    model = "Forecast";
+    model = "Forecast_Model";
     obs = {
       field = [
         {
@@ -348,7 +348,7 @@ def test_workflow__config_point_stat__sfc(c, fakefs, fmt, testvars, tidy):
         val = "ADPSFC";
       }
     ];
-    model = "Forecast";
+    model = "Forecast_Model";
     obs = {
       field = [
         {
@@ -879,7 +879,7 @@ def test_workflow__stat_reqs(baseline_name, c, statkit, cycle):
     args = (c, statkit.varname, statkit.tc, statkit.var)
     assert _stats_vs_grid.call_args_list[0].args == (
         *args,
-        f"forecast_gh_{statkit.level_type}_{statkit.level:04d}",
+        f"forecast_model_gh_{statkit.level_type}_{statkit.level:04d}",
         Source.FORECAST,
     )
 
@@ -984,7 +984,7 @@ def statkit(tc, testvars):
     return ns(
         level=level,
         level_type=level_type,
-        prefix=f"forecast_gh_{level_type}_{level:04d}",
+        prefix=f"forecast_model_gh_{level_type}_{level:04d}",
         source=Source.FORECAST,
         tc=tc,
         var=testvars[EC.gh],

src/wxvx/util.py

@@ -111,7 +111,7 @@ def classify_url(url: str) -> tuple[Proximity, str | Path]:
     if p.scheme in {"http", "https"}:
         return Proximity.REMOTE, url
     if p.scheme in {"file", ""}:
-        return Proximity.LOCAL, Path(p.path if p.scheme else url)
+        return Proximity.LOCAL, Path(p.path if p.scheme else url).resolve()
     msg = f"Scheme '{p.scheme}' in '{url}' not supported."
     raise WXVXError(msg)
 

src/wxvx/variables.py

@@ -220,6 +220,12 @@ def _levinfo(levstr: str) -> tuple[str, float | int | None]:
         return (UNKNOWN, None)
 
 
+class GDAS(GFS):
+    """
+    A GDAS variable.
+    """
+
+
 class HRRR(GFS):
     """
     A HRRR variable.

src/wxvx/workflow.py

@@ -425,7 +425,7 @@ def _netcdf_from_obs(c: Config, tc: TimeCoords):
     yield {"cfgfile": cfgfile, "prepbufr": prepbufr}
     runscript = cfgfile.ref.with_suffix(".sh")
     content = "exec time pb2nc -v 4 {prepbufr} {netcdf} {config} >{log} 2>&1".format(
-        prepbufr=prepbufr.ref,
+        prepbufr=prepbufr.ref.resolve(),
         netcdf=path,
         config=cfgfile.ref,
         log=f"{path.stem}.log",
@@ -515,7 +515,7 @@ def _stats_vs_grid(c: Config, varname: str, tc: TimeCoords, var: Var, prefix: st
     export OMP_NUM_THREADS=1
     exec time grid_stat -v 4 {fcst} {obs} {config} >{log} 2>&1
     """.format(
-        fcst=fcst.ref,
+        fcst=fcst.ref.resolve(),
         obs=obs.ref,
         config=config.ref,
         log=f"{path.stem}.log",
@@ -552,7 +552,7 @@ def _stats_vs_obs(c: Config, varname: str, tc: TimeCoords, var: Var, prefix: str
     yield reqs
     runscript = path.with_suffix(".sh")
     content = "exec time point_stat -v 4 {fcst} {obs} {config} -outdir {rundir} >{log} 2>&1".format(
-        fcst=fcst.ref,
+        fcst=fcst.ref.resolve(),
         obs=obs.ref,
         config=config.ref,
         rundir=rundir,
@@ -670,7 +670,7 @@ def _stat_args(
         cycles = c.cycles
     sections = {Source.BASELINE: c.baseline, Source.FORECAST: c.forecast, Source.TRUTH: c.truth}
     name = cast(Named, sections[source]).name.lower()
-    prefix = lambda var: "%s_%s" % (name, str(var).replace("-", "_"))
+    prefix = lambda var: "%s_%s" % (name.replace(" ", "_"), str(var).replace("-", "_"))
     args = [
         (c, vn, tc, var, prefix(var), source)
         for (var, vn), tc in product(_vxvars(c).items(), gen_timecoords(cycles, c.leadtimes))