Merge pull request #104 from bird-house/fix-output-drs-filename

Fix output drs filename

Dockerfile

@@ -1,7 +1,7 @@
 # vim:set ft=dockerfile:
 FROM continuumio/miniconda3
 MAINTAINER https://github.com/bird-house/finch
-LABEL Description="Finch WPS" Vendor="Birdhouse" Version="0.4.0"
+LABEL Description="Finch WPS" Vendor="Birdhouse" Version="0.4.1"
 
 # Update Debian system
 RUN apt-get update && apt-get install -y \

docs/source/conf.py

@@ -62,7 +62,7 @@
 # The short X.Y version.
 version = ''
 # The full version, including alpha/beta/rc tags.
-release = '0.4.0'
+release = '0.4.1'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

finch/__version__.py

@@ -6,4 +6,4 @@
 
 __author__ = """David Huard"""
 __email__ = "[email protected]"
-__version__ = "__version__ = '0.4.0'"
+__version__ = "__version__ = '0.4.1'"

finch/processes/utils.py

@@ -136,6 +136,17 @@ def compute_indices(
     output_dataset = xr.Dataset(
         data_vars=None, coords=out.coords, attrs=global_attributes
     )
+
+    # fix frequency of computed output (xclim should handle this)
+    if output_dataset.attrs.get("frequency") == "day" and "freq" in kwds:
+        conversions = {
+            "YS": "yr",
+            "MS": "mon",
+            "QS-DEC": "seasonal",
+            "AS-JUL": "seasonal",
+        }
+        output_dataset.attrs["frequency"] = conversions.get(kwds["freq"], "day")
+
     output_dataset[out.name] = out
     return output_dataset
 
@@ -155,10 +166,14 @@ def drs_filename(ds: xr.Dataset, variable: str = None):
     """
     if variable is None:
         variable = [k for k, v in ds.variables.items() if len(v.dims) >= 3][0]
+    variable = variable.replace("_", "-")
+
     # CORDEX example: tas_EUR-11_ICHEC-EC-EARTH_historical_r3i1p1_DMI-HIRHAM5_v1_day
     cordex_pattern = "{variable}_{domain}_{driving_model}_{experiment}_{ensemble}_{model}_{version}_{frequency}"
+
     # CMIP5 example: tas_MPI-ESM-LR_historical_r1i1p1
     cmip5_pattern = "{variable}_{model}_{experiment}_{ensemble}"
+
     if ds.attrs["project_id"] in ("CORDEX", "EOBS"):
         filename = cordex_pattern.format(
             variable=variable,
@@ -183,7 +198,16 @@ def drs_filename(ds: xr.Dataset, variable: str = None):
             ensemble=ensemble,
         )
     else:
-        raise Exception(f"Unknown project: {ds.attrs['project_id']}")
+        params = [
+            variable,
+            ds.attrs.get("frequency"),
+            ds.attrs.get("model_id"),
+            ds.attrs.get("driving_model_id"),
+            ds.attrs.get("experiment_id", "").replace(",", "+"),
+            ds.attrs.get("driving_experiment_id", "").replace(",", "+"),
+        ]
+        params = [k for k in params if k]
+        filename = "_".join(params)
 
     if "time" in ds:
         date_from = ds.time[0].values

setup.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.4.0
+current_version = 0.4.1
 commit = True
 tag = True
 

tests/test_utils.py

@@ -170,6 +170,15 @@ def test_drs_filename():
     assert filename == "tasmax_bcc-csm1-1_historical+rcp85_r1i1p1_19500101-19500410.nc"
 
 
+def test_drs_filename_unknown_project():
+    ds = xr.open_dataset(
+        test_data / "bccaqv2_subset_sample/tasmax_bcc-csm1-1_subset.nc"
+    )
+    ds.attrs["project_id"] = "unknown"
+    filename = drs_filename(ds)
+    assert filename == "tasmax_day_bcc-csm1-1_historical+rcp85_19500101-19500410.nc"
+
+
 def test_drs_filename_cordex():
     ds = xr.open_dataset(test_data / "cordex_subset.nc")
     filename = drs_filename(ds)

tests/test_wps_xclim_indices.py

@@ -76,7 +76,11 @@ def test_processes(client, netcdf_datasets):
         date_start = pd.to_datetime(str(ds.time[0].values))
         date_end = pd.to_datetime(str(ds.time[-1].values))
 
-        expected = f"{output_variable}_{model}_{experiment}_{ensemble}_{date_start:%Y%m%d}-{date_end:%Y%m%d}.nc"
+        expected = (
+            f"{output_variable.replace('_', '-')}_"
+            f"{model}_{experiment}_{ensemble}_"
+            f"{date_start:%Y%m%d}-{date_end:%Y%m%d}.nc"
+        )
         assert Path(outputs[0]).name == expected
 
 
@@ -128,12 +132,14 @@ def test_heat_wave_frequency_window_thresh_parameters(client, netcdf_datasets):
         wps_input_file("tasmax", netcdf_datasets["tasmax"]),
         wps_input_file("tasmin", netcdf_datasets["tasmin"]),
         wps_literal_input("window", "3"),
+        wps_literal_input("freq", "YS"),
         wps_literal_input("thresh_tasmin", "20 degC"),
         wps_literal_input("thresh_tasmax", "25 degC"),
     ]
     outputs = execute_process(client, identifier, inputs)
     ds = xr.open_dataset(outputs[0])
 
+    assert ds.attrs["frequency"] == "yr"
     assert ds.heat_wave_frequency.standard_name == _get_output_standard_name(identifier)