[MNT] RKME auto dimension

examples/examples1/example_rkme.py

@@ -3,7 +3,7 @@ import learnware.specification as specification
 
 
 if __name__ == "__main__":
-    data_X = np.random.randn(10000, 20)
+    data_X = np.random.randn(10000, 20, 10, 5)
     for i in range(10):
         data_X[i, i] = np.nan
     spec1 = specification.utils.generate_rkme_spec(X=data_X, gamma=0.1, cuda_idx=-1)
@@ -24,3 +24,5 @@ if __name__ == "__main__":
 
     print(spec1.inner_prod(spec2))
     print(spec1.dist(spec2))
+    print(spec1.get_z().shape)
+    print(spec2.get_z().shape)

learnware/specification/rkme.py

@@ -89,6 +89,9 @@ class RKMEStatSpecification(BaseStatSpecification):
         """
         alpha = None
         self.num_points = X.shape[0]
+        X_shape = X.shape
+        Z_shape = tuple([K] + list(X_shape)[1:])
+        X = X.reshape(self.num_points, -1)
 
         # fill np.nan
         X_nan = np.isnan(X)
@@ -98,7 +101,7 @@ class RKMEStatSpecification(BaseStatSpecification):
                 X[:, col] = np.where(X_nan[:, col], col_mean, X[:, col])
 
         if not reduce:
-            self.z = X
+            self.z = X.reshape(X_shape)
             self.beta = 1 / self.num_points * np.ones(self.num_points)
             self.z = torch.from_numpy(self.z).double().to(self.device)
             self.beta = torch.from_numpy(self.beta).double().to(self.device)
@@ -113,6 +116,9 @@ class RKMEStatSpecification(BaseStatSpecification):
             self._update_z(alpha, X, step_size)
             self._update_beta(X, nonnegative_beta)
 
+        # Reshape to original dimensions
+        self.z = self.z.reshape(Z_shape)
+
     def _init_z_by_faiss(self, X: Union[np.ndarray, torch.tensor], K: int):
         """Intialize Z by faiss clustering.
 
@@ -268,10 +274,10 @@ class RKMEStatSpecification(BaseStatSpecification):
         """
         beta_1 = self.beta.reshape(1, -1).double().to(self.device)
         beta_2 = Phi2.beta.reshape(1, -1).double().to(self.device)
-        Z1 = self.z.double().to(self.device)
-        Z2 = Phi2.z.double().to(self.device)
-
+        Z1 = self.z.double().reshape(self.z.shape[0], -1).to(self.device)
+        Z2 = Phi2.z.double().reshape(Phi2.z.shape[0], -1).to(self.device)        
         v = torch.sum(torch_rbf_kernel(Z1, Z2, self.gamma) * (beta_1.T @ beta_2))
+
         return float(v)
 
     def dist(self, Phi2: RKMEStatSpecification, omit_term1: bool = False) -> float:
@@ -306,6 +312,10 @@ class RKMEStatSpecification(BaseStatSpecification):
         np.ndarray
             A collection of examples which approximate the unknown distribution.
         """
+        # Flatten z
+        Z_shape = self.z.shape
+        self.z = self.z.reshape(self.z.shape[0], -1)
+        
         Nstart = 100 * T
         Xstart = self._sampling_candidates(Nstart).to(self.device)
         D = self.z[0].shape[0]
@@ -318,6 +328,11 @@ class RKMEStatSpecification(BaseStatSpecification):
             fs = (i + 1) * fsX - fsS
             idx = torch.argmax(fs)
             S[i, :] = Xstart[idx, :]
+        
+        # Reshape to orignial dimensions
+        self.z = self.z.reshape(Z_shape)
+        S_shape = tuple([S.shape[0]] + list(Z_shape)[1:])
+        S = S.reshape(S_shape)
 
         return S