Update _adaopt.py

mlsauce/adaopt/_adaopt.py

@@ -13,7 +13,9 @@
 try:
     from . import _adaoptc as adaoptc
 except ImportError:
-    import _adaoptc as adaoptc 
+    from ._adaoptc import *
+    except ImportError:
+        import _adaoptc 
 
 
 
@@ -194,7 +196,7 @@ def fit(self, X, y, **kwargs):
 
         assert n == len(y_), "must have X.shape[0] == len(y)"
 
-        res = adaoptc.fit_adaopt(
+        res = fit_adaopt(
             X=np.asarray(X_).astype(np.float64),
             y=np.asarray(y_).astype(np.int64),
             n_iterations=self.n_iterations,
@@ -268,7 +270,7 @@ def predict_proba(self, X, **kwargs):
         n_test = X.shape[0]
 
         if self.n_jobs is None:
-            return adaoptc.predict_proba_adaopt(
+            return predict_proba_adaopt(
                 X_test=np.asarray(X, order="C").astype(np.float64),
                 scaled_X_train=np.asarray(
                     self.scaled_X_train, order="C"
@@ -310,17 +312,17 @@ def multiproc_func(i):
                     p_train,
                 )
 
-                kmin_test_i = adaoptc.find_kmin_x(
+                kmin_test_i = find_kmin_x(
                     dists_test_i, n_x=n_train, k=self.k, cache=self.cache
                 )
 
-                weights_test_i = adaoptc.calculate_weights(kmin_test_i[0])
+                weights_test_i = calculate_weights(kmin_test_i[0])
 
-                probs_test_i = adaoptc.calculate_probs(
+                probs_test_i = calculate_probs(
                     kmin_test_i[1], self.probs_training
                 )
 
-                return adaoptc.average_probs(
+                return average_probs(
                     probs=probs_test_i, weights=weights_test_i
                 )
 
@@ -329,7 +331,7 @@ def multiproc_func(i):
             @delayed
             @wrap_non_picklable_objects
             def multiproc_func(i):
-                dists_test_i = adaoptc.distance_to_mat_manhattan2(
+                dists_test_i = distance_to_mat_manhattan2(
                     np.asarray(scaled_X_test.astype(np.float64), order="C")[
                         i, :
                     ],
@@ -341,17 +343,17 @@ def multiproc_func(i):
                     p_train,
                 )
 
-                kmin_test_i = adaoptc.find_kmin_x(
+                kmin_test_i = find_kmin_x(
                     dists_test_i, n_x=n_train, k=self.k, cache=self.cache
                 )
 
-                weights_test_i = adaoptc.calculate_weights(kmin_test_i[0])
+                weights_test_i = calculate_weights(kmin_test_i[0])
 
-                probs_test_i = adaoptc.calculate_probs(
+                probs_test_i = calculate_probs(
                     kmin_test_i[1], self.probs_training
                 )
 
-                return adaoptc.average_probs(
+                return average_probs(
                     probs=probs_test_i, weights=weights_test_i
                 )
 
@@ -372,17 +374,17 @@ def multiproc_func(i, *args):
                     p_train,
                 )
 
-                kmin_test_i = adaoptc.find_kmin_x(
+                kmin_test_i = find_kmin_x(
                     dists_test_i, n_x=n_train, k=self.k, cache=self.cache
                 )
 
-                weights_test_i = adaoptc.calculate_weights(kmin_test_i[0])
+                weights_test_i = calculate_weights(kmin_test_i[0])
 
-                probs_test_i = adaoptc.calculate_probs(
+                probs_test_i = calculate_probs(
                     kmin_test_i[1], self.probs_training
                 )
 
-                return adaoptc.average_probs(
+                return average_probs(
                     probs=probs_test_i, weights=weights_test_i
                 )