23059: Adds support for new analyze param - MINOR (#373)

- Adds docstrings and signature support for `use_sdm`

Author: mkoistinen

howso/client/base.py

@@ -4039,6 +4039,7 @@ def analyze(
         targeted_model: t.Optional[TargetedModel] = None,
         use_case_weights: t.Optional[bool] = None,
         use_deviations: t.Optional[bool] = None,
+        use_sdm: t.Optional[bool] = None,
         weight_feature: t.Optional[str] = None,
         **kwargs
     ):
@@ -4101,6 +4102,14 @@ def analyze(
             will be used if the Trainee has them.
         use_deviations : bool, optional
             When True, uses deviations for LK metric in queries.
+        use_sdm : bool, default None,
+            When True, Howso Engine will compute and use a sparse deviation
+            matrix (SDM) for each nominal feature in all similarity queries.
+            Enabling SDM will typically incur a small to moderate penalty on
+            speed when using nominal features in inference in exchange for
+            yielding higher quality inference. The magnitude of the changes are
+            dependent on relationships among the data and the task at hand.
+            Howso Engine will default to True, if left unset.
         weight_feature : str, optional
             Name of feature whose values to use as case weights.
             When left unspecified uses the internally managed case weight.
@@ -4165,6 +4174,7 @@ def analyze(
             rebalance_features=rebalance_features,
             targeted_model=targeted_model,
             use_deviations=use_deviations,
+            use_sdm=use_sdm,
             weight_feature=weight_feature,
         )
 
@@ -4235,6 +4245,7 @@ def set_auto_analyze_params(
         targeted_model: t.Optional[TargetedModel] = None,
         use_deviations: t.Optional[bool] = None,
         use_case_weights: t.Optional[bool] = None,
+        use_sdm: t.Optional[bool] = None,
         weight_feature: t.Optional[str] = None,
         **kwargs
     ):
@@ -4307,6 +4318,14 @@ def set_auto_analyze_params(
             If set to True, will scale influence weights by each case's
             `weight_feature` weight. If unspecified, case weights
             will be used if the Trainee has them.
+        use_sdm : bool, default None,
+            When True, Howso Engine will compute and use a sparse deviation
+            matrix (SDM) for each nominal feature in all similarity queries.
+            Enabling SDM will typically incur a small to moderate penalty on
+            speed when using nominal features in inference in exchange for
+            yielding higher quality inference. The magnitude of the changes are
+            dependent on relationships among the data and the task at hand.
+            Howso Engine will default to True, if left unset.
         weight_feature : str
             Name of feature whose values to use as case weights.
             When left unspecified uses the internally managed case weight.
@@ -4395,6 +4414,7 @@ def set_auto_analyze_params(
             "use_deviations": use_deviations,
             "inverse_residuals_as_weights": inverse_residuals_as_weights,
             "use_case_weights": use_case_weights,
+            "use_sdm": use_sdm,
             "weight_feature": weight_feature,
             **kwargs,
         })

howso/engine/trainee.py

@@ -989,6 +989,7 @@ def analyze(
         targeted_model: t.Optional[TargetedModel] = None,
         use_case_weights: t.Optional[bool] = None,
         use_deviations: t.Optional[bool] = None,
+        use_sdm: t.Optional[bool] = True,
         weight_feature: t.Optional[str] = None,
         **kwargs
     ):
@@ -1049,6 +1050,14 @@ def analyze(
             be used if the Trainee has them.
         use_deviations : bool, default False
             When True, uses deviations for LK metric in queries.
+        use_sdm : bool, default None,
+            When True, Howso Engine will compute and use a sparse deviation
+            matrix (SDM) for each nominal feature in all similarity queries.
+            Enabling SDM will typically incur a small to moderate penalty on
+            speed when using nominal features in inference in exchange for
+            yielding higher quality inference. The magnitude of the changes are
+            dependent on relationships among the data and the task at hand.
+            Howso Engine will default to True, if left unset.
         weight_feature : str, optional
             Name of feature whose values to use as case weights.
             When left unspecified uses the internally managed case weight.
@@ -1075,6 +1084,7 @@ def analyze(
                 rebalance_features=rebalance_features,
                 targeted_model=targeted_model,
                 use_deviations=use_deviations,
+                use_sdm=use_sdm,
                 weight_feature=weight_feature,
                 **kwargs
             )

version.json

@@ -1,5 +1,5 @@
 {
   "dependencies": {
-    "howso-engine": "97.1.1"
+    "howso-engine": "97.2.0"
   }
 }