dont pass sample fraction at predict (#62)

crflynn · Jan 18, 2021 · d90612d · d90612d
1 parent d409bd8
commit d90612d
Show file tree

Hide file tree

Showing 6 changed files with 48 additions and 4 deletions.
diff --git a/skranger/ensemble/ranger_forest_classifier.py b/skranger/ensemble/ranger_forest_classifier.py
@@ -269,7 +269,7 @@ def predict_proba(self, X):
             self.class_weights or [],
             False,  # predict_all
             self.keep_inbag,
-            self.sample_fraction_,
+            [1],  # sample_fraction
             0.5,  # alpha
             0.1,  # minprop
             self.holdout,

diff --git a/skranger/ensemble/ranger_forest_regressor.py b/skranger/ensemble/ranger_forest_regressor.py
@@ -272,7 +272,7 @@ def _get_terminal_node_forest(self, X):
             [],  # class_weights
             False,  # predict_all
             self.keep_inbag,
-            self.sample_fraction_,
+            [1],  # sample_fraction
             0,  # alpha
             0,  # minprop
             self.holdout,
@@ -354,7 +354,7 @@ def predict(self, X):
             [],  # class_weights
             False,  # predict_all
             self.keep_inbag,
-            self.sample_fraction_,
+            [1],  # sample_fraction
             self.alpha,
             self.minprop,
             self.holdout,

diff --git a/skranger/ensemble/ranger_forest_survival.py b/skranger/ensemble/ranger_forest_survival.py
@@ -255,7 +255,7 @@ def _predict(self, X):
             [],  # class_weights
             False,  # predict_all
             self.keep_inbag,
-            self.sample_fraction_,
+            [1],  # sample_fraction
             self.alpha,
             self.minprop,
             self.holdout,

diff --git a/tests/ensemble/test_ranger_forest_classifier.py b/tests/ensemble/test_ranger_forest_classifier.py
@@ -36,18 +36,33 @@ def test_predict(self, iris_X, iris_y):
         pred = rfc.predict(iris_X)
         assert len(pred) == iris_X.shape[0]
 
+        # test with single record
+        iris_X_record = iris_X[0:1, :]
+        pred = rfc.predict(iris_X_record)
+        assert len(pred) == 1
+
     def test_predict_proba(self, iris_X, iris_y):
         rfc = RangerForestClassifier()
         rfc.fit(iris_X, iris_y)
         pred = rfc.predict_proba(iris_X)
         assert len(pred) == iris_X.shape[0]
 
+        # test with single record
+        iris_X_record = iris_X[0:1, :]
+        pred = rfc.predict_proba(iris_X_record)
+        assert len(pred) == 1
+
     def test_predict_log_proba(self, iris_X, iris_y):
         rfc = RangerForestClassifier()
         rfc.fit(iris_X, iris_y)
         pred = rfc.predict_log_proba(iris_X)
         assert len(pred) == iris_X.shape[0]
 
+        # test with single record
+        iris_X_record = iris_X[0:1, :]
+        pred = rfc.predict_log_proba(iris_X_record)
+        assert len(pred) == 1
+
     def test_serialize(self, iris_X, iris_y):
         tf = tempfile.TemporaryFile()
         rfc = RangerForestClassifier()
@@ -145,6 +160,15 @@ def test_sample_fraction(self, iris_X, iris_y):
         rfc.fit(iris_X, iris_y)
         assert rfc.sample_fraction_ == [0.69]
 
+        # test with single record
+        iris_X_record = iris_X[0:1, :]
+        pred = rfc.predict(iris_X_record)
+        assert len(pred) == 1
+        pred = rfc.predict_proba(iris_X_record)
+        assert len(pred) == 1
+        pred = rfc.predict_log_proba(iris_X_record)
+        assert len(pred) == 1
+
     def test_sample_fraction_replace(self, iris_X, iris_y, replace):
         rfc = RangerForestClassifier(replace=replace)
         rfc.fit(iris_X, iris_y)

diff --git a/tests/ensemble/test_ranger_forest_regressor.py b/tests/ensemble/test_ranger_forest_regressor.py
@@ -31,6 +31,11 @@ def test_predict(self, boston_X, boston_y):
         pred = rfr.predict(boston_X)
         assert len(pred) == boston_X.shape[0]
 
+        # test with single record
+        boston_X_record = boston_X[0:1, :]
+        pred = rfr.predict(boston_X_record)
+        assert len(pred) == 1
+
     def test_serialize(self, boston_X, boston_y):
         tf = tempfile.TemporaryFile()
         rfr = RangerForestRegressor()
@@ -125,6 +130,11 @@ def test_sample_fraction(self, iris_X, iris_y):
         rfr.fit(iris_X, iris_y)
         assert rfr.sample_fraction_ == [0.69]
 
+        # test with single record
+        iris_X_record = iris_X[0:1, :]
+        pred = rfr.predict(iris_X_record)
+        assert len(pred) == 1
+
     def test_sample_fraction_replace(self, boston_X, boston_y, replace):
         rfr = RangerForestRegressor(replace=replace)
         rfr.fit(boston_X, boston_y)

diff --git a/tests/ensemble/test_ranger_forest_survival.py b/tests/ensemble/test_ranger_forest_survival.py
@@ -34,6 +34,11 @@ def test_predict(self, lung_X, lung_y):
         pred = rfs.predict(lung_X)
         assert len(pred) == lung_X.shape[0]
 
+        # test with single record
+        lung_X_record = lung_X.values[0:1, :]
+        pred = rfs.predict(lung_X_record)
+        assert len(pred) == 1
+
     def test_predict_cumulative_hazard_function(self, lung_X, lung_y):
         rfs = RangerForestSurvival(n_estimators=N_ESTIMATORS)
         rfs.fit(lung_X, lung_y)
@@ -140,6 +145,11 @@ def test_sample_fraction(self, lung_X, lung_y):
         rfs.fit(lung_X, lung_y)
         assert rfs.sample_fraction_ == [0.69]
 
+        # test with single record
+        lung_X_record = lung_X.values[0:1, :]
+        pred = rfs.predict(lung_X_record)
+        assert len(pred) == 1
+
     def test_sample_fraction_replace(self, lung_X, lung_y, replace):
         rfs = RangerForestSurvival(replace=replace)
         rfs.fit(lung_X, lung_y)