Make AbstractBPCache an AbstractITensorNetwork

Author: JoeyT1994

Project.toml

@@ -1,7 +1,7 @@
 name = "ITensorNetworks"
 uuid = "2919e153-833c-4bdc-8836-1ea460a35fc7"
 authors = ["Matthew Fishman <[email protected]>, Joseph Tindall <[email protected]> and contributors"]
-version = "0.13.7"
+version = "0.13.8"
 
 [deps]
 AbstractTrees = "1520ce14-60c1-5f80-bbc7-55ef81b5835c"

src/caches/abstractbeliefpropagationcache.jl

@@ -11,9 +11,18 @@ using NamedGraphs.PartitionedGraphs:
   partitionedges,
   unpartitioned_graph
 using SimpleTraits: SimpleTraits, Not, @traitfn
+using NamedGraphs.SimilarType: SimilarType
 using NDTensors: NDTensors
 
-abstract type AbstractBeliefPropagationCache end
+abstract type AbstractBeliefPropagationCache{V,PV} <: AbstractITensorNetwork{V} end
+
+function SimilarType.similar_type(bpc::AbstractBeliefPropagationCache)
+  return typeof(tensornetwork(bpc))
+end
+function data_graph_type(bpc::AbstractBeliefPropagationCache)
+  return data_graph_type(tensornetwork(bpc))
+end
+data_graph(bpc::AbstractBeliefPropagationCache) = data_graph(tensornetwork(bpc))
 
 function default_message_update(contract_list::Vector{ITensor}; normalize=true, kwargs...)
   sequence = contraction_sequence(contract_list; alg="optimal")
@@ -40,6 +49,9 @@ default_messages(ptn::PartitionedGraph) = Dictionary()
 end
 default_partitioned_vertices(ψ::AbstractITensorNetwork) = group(v -> v, vertices(ψ))
 
+function Base.setindex!(bpc::AbstractBeliefPropagationCache, factor::ITensor, vertex)
+  not_implemented()
+end
 partitioned_tensornetwork(bpc::AbstractBeliefPropagationCache) = not_implemented()
 messages(bpc::AbstractBeliefPropagationCache) = not_implemented()
 function default_message(
@@ -88,12 +100,8 @@ function tensornetwork(bpc::AbstractBeliefPropagationCache)
   return unpartitioned_graph(partitioned_tensornetwork(bpc))
 end
 
-function setindex_preserve_graph!(bpc::AbstractBeliefPropagationCache, args...)
-  return setindex_preserve_graph!(tensornetwork(bpc), args...)
-end
-
 function factors(bpc::AbstractBeliefPropagationCache, verts::Vector)
-  return ITensor[tensornetwork(bpc)[v] for v in verts]
+  return ITensor[copy(bpc[v]) for v in verts]
 end
 
 function factors(
@@ -143,7 +151,6 @@ for f in [
   :(PartitionedGraphs.partitionvertices),
   :(PartitionedGraphs.vertices),
   :(PartitionedGraphs.boundary_partitionedges),
-  :(linkinds),
 ]
   @eval begin
     function $f(bpc::AbstractBeliefPropagationCache, args...; kwargs...)
@@ -152,23 +159,28 @@ for f in [
   end
 end
 
+function linkinds(bpc::AbstractBeliefPropagationCache, pe::PartitionEdge)
+  return linkinds(partitioned_tensornetwork(bpc), pe)
+end
+
 NDTensors.scalartype(bpc::AbstractBeliefPropagationCache) = scalartype(tensornetwork(bpc))
 
 """
-Update the tensornetwork inside the cache
+Update the tensornetwork inside the cache out-of-place
 """
 function update_factors(bpc::AbstractBeliefPropagationCache, factors)
   bpc = copy(bpc)
-  tn = tensornetwork(bpc)
   for vertex in eachindex(factors)
     # TODO: Add a check that this preserves the graph structure.
-    setindex_preserve_graph!(tn, factors[vertex], vertex)
+    setindex_preserve_graph!(bpc, factors[vertex], vertex)
   end
   return bpc
 end
 
 function update_factor(bpc, vertex, factor)
-  return update_factors(bpc, Dictionary([vertex], [factor]))
+  bpc = copy(bpc)
+  setindex_preserve_graph!(bpc, factor, vertex)
+  return bpc
 end
 
 function message(bpc::AbstractBeliefPropagationCache, edge::PartitionEdge; kwargs...)
@@ -178,12 +190,45 @@ end
 function messages(bpc::AbstractBeliefPropagationCache, edges; kwargs...)
   return map(edge -> message(bpc, edge; kwargs...), edges)
 end
+function set_messages!(bpc::AbstractBeliefPropagationCache, partitionedges_messages)
+  ms = messages(bpc)
+  for pe in eachindex(partitionedges_messages)
+    # TODO: Add a check that this preserves the graph structure.
+    set!(ms, pe, partitionedges_messages[pe])
+  end
+  return bpc
+end
+function set_message!(bpc::AbstractBeliefPropagationCache, pe::PartitionEdge, message)
+  ms = messages(bpc)
+  set!(ms, pe, message)
+  return bpc
+end
+
+function set_messages(bpc::AbstractBeliefPropagationCache, partitionedges_messages)
+  bpc = copy(bpc)
+  return set_messages!(bpc, partitionedges_messages)
+end
 function set_message(bpc::AbstractBeliefPropagationCache, pe::PartitionEdge, message)
   bpc = copy(bpc)
+  return set_message!(bpc, pe, message)
+end
+function delete_messages!(bpc::AbstractBeliefPropagationCache, pes::Vector{<:PartitionEdge})
   ms = messages(bpc)
-  set!(ms, pe, message)
+  for pe in pes
+    delete!(ms, pe)
+  end
   return bpc
 end
+function delete_message!(bpc::AbstractBeliefPropagationCache, pe::PartitionEdge)
+  return delete_message!(bpc, [pe])
+end
+function delete_messages(bpc::AbstractBeliefPropagationCache, pes::Vector{<:PartitionEdge})
+  bpc = copy(bpc)
+  return delete_messages!(bpc, pes)
+end
+function delete_message(bpc::AbstractBeliefPropagationCache, pe::PartitionEdge)
+  return delete_message(bpc, [pe])
+end
 
 """
 Compute message tensor as product of incoming mts and local state
@@ -241,11 +286,11 @@ function update(
   edge_groups::Vector{<:Vector{<:PartitionEdge}};
   kwargs...,
 )
-  new_mts = copy(messages(bpc))
+  new_mts = Dictionary()
   for edges in edge_groups
     bpc_t = update(alg, bpc, edges; kwargs...)
     for e in edges
-      new_mts[e] = message(bpc_t, e)
+      set!(new_mts, e, message(bpc_t, e))
     end
   end
   return set_messages(bpc, new_mts)
@@ -288,10 +333,6 @@ function update(
   return update(Algorithm(alg), bpc; kwargs...)
 end
 
-function scale!(bp_cache::AbstractBeliefPropagationCache, args...)
-  return scale!(tensornetwork(bp_cache), args...)
-end
-
 function rescale_messages(
   bp_cache::AbstractBeliefPropagationCache, partitionedge::PartitionEdge
 )

src/caches/beliefpropagationcache.jl

@@ -3,6 +3,7 @@ using SplitApplyCombine: group
 using LinearAlgebra: diag, dot
 using ITensors: dir
 using NamedGraphs.PartitionedGraphs:
+  AbstractPartitionedGraph,
   PartitionedGraphs,
   PartitionedGraph,
   PartitionVertex,
@@ -23,7 +24,8 @@ function default_cache_construction_kwargs(alg::Algorithm"bp", pg::PartitionedGr
   return (;)
 end
 
-struct BeliefPropagationCache{PTN,MTS} <: AbstractBeliefPropagationCache
+struct BeliefPropagationCache{V,PV,PTN<:AbstractPartitionedGraph{V,PV},MTS} <:
+       AbstractBeliefPropagationCache{V,PV}
   partitioned_tensornetwork::PTN
   messages::MTS
 end
@@ -81,15 +83,12 @@ function default_message_update_kwargs(
   return (;)
 end
 
+Base.setindex!(bpc::BeliefPropagationCache, factor::ITensor, vertex) = not_implemented()
 partitions(bpc::BeliefPropagationCache) = partitionvertices(partitioned_tensornetwork(bpc))
 function PartitionedGraphs.partitionedges(bpc::BeliefPropagationCache)
   partitionedges(partitioned_tensornetwork(bpc))
 end
 
-function set_messages(cache::BeliefPropagationCache, messages)
-  return BeliefPropagationCache(partitioned_tensornetwork(cache), messages)
-end
-
 function environment(bpc::BeliefPropagationCache, verts::Vector; kwargs...)
   partition_verts = partitionvertices(bpc, verts)
   messages = incoming_messages(bpc, partition_verts; kwargs...)

test/test_belief_propagation.jl

@@ -6,6 +6,7 @@ using ITensorNetworks:
   ITensorNetworks,
   BeliefPropagationCache,
   ⊗,
+  @preserve_graph,
   combine_linkinds,
   contract,
   contraction_sequence,
@@ -48,6 +49,19 @@ using Test: @test, @testset
     ψ = random_tensornetwork(rng, elt, s; link_space=χ)
     ψψ = ψ ⊗ prime(dag(ψ); sites=[])
     bpc = BeliefPropagationCache(ψψ, group(v -> first(v), vertices(ψψ)))
+
+    #Test updating the tensors in the cache
+    vket, vbra = ((1, 1), 1), ((1, 1), 2)
+    A = bpc[vket]
+    new_A = random_itensor(elt, inds(A))
+    new_A_dag = ITensors.replaceind(
+      dag(prime(new_A)), only(s[first(vket)])', only(s[first(vket)])
+    )
+    @preserve_graph bpc[vket] = new_A
+    @preserve_graph bpc[vbra] = new_A_dag
+    @test bpc[vket] == new_A
+    @test bpc[vbra] == new_A_dag
+
     bpc = update(bpc; maxiter=25, tol=eps(real(elt)))
     #Test messages are converged
     for pe in partitionedges(bpc)