Check that we can pass feature_names and coef_scale directly

Also add coef_scale to explain_prediction kwargs
TeamHG-Memex · Sep 26, 2016 · 6da8bb3 · 6da8bb3
1 parent 55af1c8
commit 6da8bb3
Show file tree

Hide file tree

Showing 3 changed files with 8 additions and 4 deletions.
diff --git a/eli5/sklearn/explain_prediction.py b/eli5/sklearn/explain_prediction.py
@@ -37,7 +37,7 @@
 
 @singledispatch
 def explain_prediction(clf, doc, vec=None, top=_TOP, target_names=None,
-                       feature_names=None, vectorized=False):
+                       feature_names=None, vectorized=False, coef_scale=None):
     """ Return an explanation of an estimator """
     return {
         "estimator": repr(clf),

diff --git a/eli5/sklearn/explain_weights.py b/eli5/sklearn/explain_weights.py
@@ -149,7 +149,7 @@ def explain_linear_classifier_weights(clf, vec=None, top=_TOP, target_names=None
     To print it use utilities from eli5.formatters.
     """
     feature_names, coef_scale = handle_hashing_vec(vec, feature_names,
-                                                    coef_scale)
+                                                   coef_scale)
     feature_names = get_feature_names(clf, vec, feature_names=feature_names)
     _extra_caveats = "\n" + HASHING_CAVEATS if is_invhashing(vec) else ''
 

diff --git a/tests/test_sklearn_explain_prediction.py b/tests/test_sklearn_explain_prediction.py
@@ -115,8 +115,8 @@ def test_explain_hashing_vectorizer(newsgroups_train_binary):
     X = vec.fit_transform(docs)
     clf.fit(X, y)
 
-    get_res = lambda: explain_prediction(
-        clf, docs[0], ivec, target_names=target_names, top=20)
+    get_res = lambda **kwargs: explain_prediction(
+        clf, docs[0], ivec, target_names=target_names, top=20, **kwargs)
     res = get_res()
     check_explain_linear_binary(res)
     assert res == get_res()
@@ -126,6 +126,10 @@ def test_explain_hashing_vectorizer(newsgroups_train_binary):
     pprint(res_vectorized)
     assert res_vectorized == res
 
+    assert res == get_res(
+        feature_names=ivec.get_feature_names(always_signed=False),
+        coef_scale=ivec.column_signs_)
+
 
 def test_explain_linear_dense():
     clf = LogisticRegression()