Group metrics options in their own config section

docs/more_complete_config_example.toml

@@ -46,14 +46,20 @@
     # Note that the interval is closed at the left and open at the right.
     dtgs.cycle_length = '3H' # Default = '3H'
     #
-    # custom_metrics_optimize_mode:
+
+[metrics]
+    # method:
+    #    > default: "haversine"
+    #    > choices: "correlation", "haversine"
+    method = "correlation"
+    # optimize_mode:
     #    > default: "memory"
     #    > choices: "memory", "speed_mem_compromise", "speed"
     # In terms of memory usage:
     #     "memory" < "speed_mem_compromise" < "speed"
     # In terms of execution time:
     #     "memory" > "speed_mem_compromise" > "speed"
-    custom_metrics_optimize_mode = "memory"
+    optimize_mode = "memory"
 
 ##################################
 # Options controlling the domain #

netatmoqc/apps/clustering/app.py

@@ -660,13 +660,12 @@ def run_clustering_and_make_plot(
         {
             "general": dict(
                 clustering_method=method,
-                custom_metrics_optimize_mode=config.general.custom_metrics_optimize_mode,
-                custom_metrics_method=config.general.custom_metrics_method,
                 dtgs=dict(
                     list=config.general.dtgs,
                     cycle_length=config.general.dtgs.cycle_length.freqstr,
                 ),
             ),
+            "metrics": config.metrics,
             "clustering_method.%s"
             % (method): dict(
                 eps=eps,

netatmoqc/config_parser.py

@@ -379,19 +379,6 @@ with config_section("general") as section:
     config_metadata.register("dtgs.start")
     config_metadata.register("dtgs.end")
     config_metadata.register("dtgs.cycle_length", default="3H")
-    # TODO: Group metrics-related options in their own config section
-    # Metrics optimisaion scheme
-    config_metadata.register(
-        "custom_metrics_optimize_mode",
-        default="memory",
-        choices=["speed", "memory", "speed_mem_compromise"],
-    )
-    # Metrics calculation method
-    config_metadata.register(
-        "custom_metrics_method",
-        default="haversine",
-        choices=["correlation", "haversine"],
-    )
     # Data cols to ignore when running clustering
     config_metadata.register(
         "unclusterable_data_columns", default=["id", "time_utc"]
@@ -402,7 +389,20 @@ with config_section("general") as section:
         default=["pressure", "temperature", "humidity", "sum_rain_1"],
     )
 
-# "commans section"
+# "metrics" section
+with config_section("metrics") as section:
+    config_metadata.register(
+        "method",
+        default="haversine",
+        choices=["correlation", "haversine"],
+    )
+    config_metadata.register(
+        "optimize_mode",
+        default="memory",
+        choices=["speed", "memory", "speed_mem_compromise"],
+    )
+
+# "commands" section
 with config_section("commands.select") as section:
     config_metadata.register("station_rejection_tol", default=0.15, minval=0.0)
 

netatmoqc/metrics.py

@@ -331,7 +331,7 @@ def calc_distance_matrix(df, config, domain=None, num_threads=-1):
 
     """
     accepted_methods = ["correlation", "haversine"]
-    method = config.general.custom_metrics_method.lower()
+    method = config.metrics.method.lower()
     if method not in accepted_methods:
         raise NotImplementedError(
             "Distance matrix calc method '%s' not available. " % (method)
@@ -346,7 +346,7 @@ def calc_distance_matrix(df, config, domain=None, num_threads=-1):
             # Drop columns that won't be used in the clustering
             df=df,
             weights=weights_dict_to_np_array(df, config=config),
-            optimize_mode=config.general.custom_metrics_optimize_mode,
+            optimize_mode=config.metrics.optimize_mode,
             num_threads=num_threads,
         )
 
@@ -360,7 +360,7 @@ def calc_distance_matrix(df, config, domain=None, num_threads=-1):
         return calc_distance_matrix_considering_correlation(
             df=df,
             weights_dict=weights_dict,
-            optimize_mode=config.general.custom_metrics_optimize_mode,
+            optimize_mode=config.metrics.optimize_mode,
             num_threads=num_threads,
             domain=domain,
         )

tests/test_metrics.py

@@ -40,10 +40,10 @@ class TestsCustomMetrics:
         # Prepare a mock config
         fpath = Path(__file__).resolve().parent / "test_config.toml"
         config = ParsedConfig(fpath)
-        config.general.custom_metrics_method = "haversine"
+        config.metrics.method = "haversine"
         config.general.clustering_method = "hdbscan"
         config.clustering_method.hdbscan.obs_weights = {"some_obs": 0.0}
-        config.general.custom_metrics_optimize_mode = "speed"
+        config.metrics.optimize_mode = "speed"
         config.general.unclusterable_data_columns = []
 
         logger.info("Calculating %dx%d distance matrix:", n_pts, n_pts)