add tensor_1d_compress.py

quimb/tensor/__init__.py

@@ -129,6 +129,7 @@
     NNI_ham_XY,
     SpinHam,
     SpinHam1D,
+    TN1D_matching,
     TN2D_classical_ising_partition_function,
     TN2D_corner_double_line,
     TN2D_embedded_classical_ising_partition_function,
@@ -361,6 +362,7 @@
     "TN_rand_from_edges",
     "TN_rand_reg",
     "TN_rand_tree",
+    "TN1D_matching",
     "TN2D_classical_ising_partition_function",
     "TN2D_corner_double_line",
     "TN2D_embedded_classical_ising_partition_function",

quimb/tensor/tensor_1d.py

@@ -1,5 +1,4 @@
-"""Classes and algorithms related to 1D tensor networks.
-"""
+"""Classes and algorithms related to 1D tensor networks."""
 
 import operator
 import functools
@@ -80,27 +79,15 @@ def expec_TN_1D(*tns, compress=None, eps=1e-15):
     return expec_tn ^ ...
 
 
-_VALID_GATE_CONTRACT = {
-    False,
-    True,
-    "swap+split",
-    "split-gate",
-    "swap-split-gate",
-    "auto-split-gate",
-}
-_VALID_GATE_PROPAGATE = {"sites", "register", False, True}
-_TWO_BODY_ONLY = _VALID_GATE_CONTRACT - {True, False}
-
-
-def maybe_factor_gate_into_tensor(G, dp, ng, where):
+def maybe_factor_gate_into_tensor(G, phys_dim, nsites, where):
     # allow gate to be a matrix as long as it factorizes into tensor
-    shape_matches_2d = (ops.ndim(G) == 2) and (G.shape[1] == dp**ng)
-    shape_matches_nd = all(d == dp for d in G.shape)
+    shape_matches_2d = (ops.ndim(G) == 2) and (G.shape[1] == phys_dim**nsites)
+    shape_matches_nd = all(d == phys_dim for d in G.shape)
 
     if shape_matches_2d:
         G = ops.asarray(G)
-        if ng >= 2:
-            G = reshape(G, [dp] * 2 * ng)
+        if nsites >= 2:
+            G = reshape(G, [phys_dim] * 2 * nsites)
 
     elif not shape_matches_nd:
         raise ValueError(
@@ -1126,8 +1113,7 @@ def compress_site(
             self.right_compress_site(i + 1, bra=bra, **compress_opts)
 
     def bond(self, i, j):
-        """Get the name of the index defining the bond between sites i and j.
-        """
+        """Get the name of the index defining the bond between sites i and j."""
         (bond,) = self[i].bonds(self[j])
         return bond
 
@@ -1481,7 +1467,12 @@ def from_fill_fn(
         else:
             phys_dims = itertools.cycle(phys_dim)
 
-        mps = TensorNetwork()
+        mps = cls.new(
+            L=L,
+            cyclic=cyclic,
+            site_ind_id=site_ind_id,
+            site_tag_id=site_tag_id,
+        )
         global_tags = tags_to_oset(tags)
         bonds = collections.defaultdict(rand_uuid)
 
@@ -1504,13 +1495,7 @@ def from_fill_fn(
             tags = global_tags | oset((site_tag_id.format(i),))
             mps |= Tensor(data, inds=inds, tags=tags)
 
-        return mps.view_as_(
-            cls,
-            L=L,
-            cyclic=cyclic,
-            site_ind_id=site_ind_id,
-            site_tag_id=site_tag_id,
-        )
+        return mps
 
     @classmethod
     def from_dense(
@@ -2887,6 +2872,65 @@ def gen_tensors():
 
         super().__init__(gen_tensors(), virtual=True, **tn_opts)
 
+    @classmethod
+    def from_fill_fn(
+        cls,
+        fill_fn,
+        L,
+        bond_dim,
+        phys_dim=2,
+        cyclic=False,
+        shape="lrud",
+        site_tag_id="I{}",
+        tags=None,
+        upper_ind_id="k{}",
+        lower_ind_id="b{}",
+    ):
+        if set(shape) - {"l", "r", "u", "d"}:
+            raise ValueError(f"Invalid shape string: {shape}.")
+
+        # check for site varying physical dimensions
+        if isinstance(phys_dim, Integral):
+            phys_dims = itertools.repeat(phys_dim)
+        else:
+            phys_dims = itertools.cycle(phys_dim)
+
+        mpo = cls.new(
+            L=L,
+            cyclic=cyclic,
+            site_tag_id=site_tag_id,
+            upper_ind_id=upper_ind_id,
+            lower_ind_id=lower_ind_id,
+        )
+
+        global_tags = tags_to_oset(tags)
+        bonds = collections.defaultdict(rand_uuid)
+
+        for i in range(L):
+            p = next(phys_dims)
+            inds = []
+            data_shape = []
+            for c in shape:
+                if c == "l":
+                    if (i - 1) >= 0 or cyclic:
+                        inds.append(bonds[frozenset([(i - 1) % L, i])])
+                        data_shape.append(bond_dim)
+                elif c == "r":
+                    if (i + 1) < L or cyclic:
+                        inds.append(bonds[frozenset([i, (i + 1) % L])])
+                        data_shape.append(bond_dim)
+                elif c == "u":
+                    inds.append(upper_ind_id.format(i))
+                    data_shape.append(p)
+                else:  # c == "d"
+                    inds.append(lower_ind_id.format(i))
+                    data_shape.append(p)
+            data = fill_fn(data_shape)
+            tags = global_tags | oset((site_tag_id.format(i),))
+            mpo |= Tensor(data, inds=inds, tags=tags)
+
+        return mpo
+
     def add_MPO(self, other, inplace=False, compress=False, **compress_opts):
         """Add another MatrixProductState to this one."""
         if self.L != other.L: