Working on rbf memory issue.

analysistools/reweighting/rbf_reweighter.py

@@ -96,7 +96,20 @@ class RbfReweighter:
         #       and
         #       https://numpy.org/doc/stable/reference/generated/numpy.hsplit.html
 
-        weights = self._rbf(*[origin_sample[variable].values for variable in self._variables])
+        weights = np.zeros(shape=len(origin_sample.index), dtype=np.float64)
+        for arrays in np.nditer(
+                [weights] + [origin_sample[v].values for v in self._variables],
+                flags=['external_loop', 'buffered'],
+                op_flags=['readwrite']
+        ):
+            assert len(arrays) == 1 + len(self._variables), (len(arrays), len(self._variables))
+            i_weights = arrays[0]
+            assert all(isinstance(a, np.ndarray) for a in arrays), [type(a) for a in arrays]
+            assert all(a.shape == i_weights.shape for a in arrays), [a.shape for a in arrays]
+            i_weights = self._rbf(*arrays[1:])
+            assert len(i_weights.shape) == 1, i_weights.shape
+
+        # weights = self._rbf(*[origin_sample[variable].values for variable in self._variables])
         weights[weights < 0] = 0.0
         num_origin_events = len(origin_sample.index)
         summed_weights = np.sum(weights)
@@ -108,11 +121,16 @@ class RbfReweighter:
 
         logging.debug(f"Scale factor: {self._scale_factor}")
 
+    @staticmethod
+    def euclidean_norm_numpy(x1, x2):
+        return np.linalg.norm(x1 - x2, axis=0)
+
     def _create_rbf(self):
         self._rbf = Rbf(
             *[self._weight_coords[:, i] for i in range(len(self._variables))],
             self._hist_weights,
-            function="cubic"
+            function="cubic",
+            norm=RbfReweighter.euclidean_norm_numpy
         )
 
     def export_to_json(self, output_path: Union[AnyStr, os.PathLike]):