Merge pull request #46 from guitarmind/master

Fix TypeError: iteration over a 0-d array error when no features are rejected

boruta/boruta_py.py

@@ -331,20 +331,24 @@ def _fit(self, X, y):
         not_selected = np.setdiff1d(np.arange(n_feat), selected)
         # large importance values should rank higher = lower ranks -> *(-1)
         imp_history_rejected = imp_history[1:, not_selected] * -1
-        # calculate ranks in each iteration, then median of ranks across feats
-        iter_ranks = self._nanrankdata(imp_history_rejected, axis=1)
-        rank_medians = np.nanmedian(iter_ranks, axis=0)
-        ranks = self._nanrankdata(rank_medians, axis=0)
 
         # update rank for not_selected features
-        if not_selected.shape[0] > 0:
-            # set smallest rank to 3 if there are tentative feats
-            if tentative.shape[0] > 0:
-                ranks = ranks - np.min(ranks) + 3
-            else:
-                # and 2 otherwise
-                ranks = ranks - np.min(ranks) + 2
-            self.ranking_[not_selected] = ranks
+        if not_selected.shape[0] > 0 and not_selected.shape[1] > 0:
+                # calculate ranks in each iteration, then median of ranks across feats
+                iter_ranks = self._nanrankdata(imp_history_rejected, axis=1)
+                rank_medians = np.nanmedian(iter_ranks, axis=0)
+                ranks = self._nanrankdata(rank_medians, axis=0)
+
+                # set smallest rank to 3 if there are tentative feats
+                if tentative.shape[0] > 0:
+                    ranks = ranks - np.min(ranks) + 3
+                else:
+                    # and 2 otherwise
+                    ranks = ranks - np.min(ranks) + 2
+                self.ranking_[not_selected] = ranks
+        else:
+            # all are selected, thus we set feature supports to True
+            self.support_ = np.ones(n_feat, dtype=np.bool)
 
         # notify user
         if self.verbose > 0: