Rename SimpleTensorNetwork => TensorNetwork

Author: shinaoka

src/simpletensornetwork.jl

@@ -1,24 +1,23 @@
 """
 Generic tensor network data structure
 """
-mutable struct SimpleTensorNetwork <: AbstractDataGraph{Int,IndexedArray,IndexedArray}
+mutable struct TensorNetwork <: AbstractDataGraph{Int,IndexedArray,IndexedArray}
     # data_graph: (undirected) graph of the tensor network
     #   An integer is assigned to each vertex (starting from 1 and increasing one by one).
     #   We can place an IndexedArray at each vertex of the graph, and an edge between two vertices.
     #   But, the latter is not supported by the current implementation of SimpleTensorNetworks.jl.
     #   This may be useful for supporting the Vidal notation.
     data_graph::DataGraph{Int,IndexedArray,IndexedArray,NamedGraph{Int},NamedEdge{Int}}
 
-    function SimpleTensorNetwork(
+    function TensorNetwork(
         dg::DataGraph{Int,IndexedArray,IndexedArray,NamedGraph{Int},NamedEdge{Int}},
     )
-        is_connected(dg) ||
-            error("SimpleTensorNetwork is only supported for a connected graph.")
+        is_connected(dg) || error("TensorNetwork is only supported for a connected graph.")
         new(dg)
     end
 end
 
-function SimpleTensorNetwork(ts::AbstractVector{<:AbstractIndexedArray})
+function TensorNetwork(ts::AbstractVector{<:AbstractIndexedArray})
     g = NamedGraph(collect(eachindex(ts)))
     dg = DataGraph{Int,IndexedArray,IndexedArray,NamedGraph{Int},NamedEdge{Int}}(g)
 
@@ -32,34 +31,33 @@ function SimpleTensorNetwork(ts::AbstractVector{<:AbstractIndexedArray})
             end
         end
     end
-    tn = SimpleTensorNetwork(dg)
+    tn = TensorNetwork(dg)
     return tn
 end
 
-data_graph(tn::SimpleTensorNetwork) = getfield(tn, :data_graph)
-data_graph_type(TN::Type{<:SimpleTensorNetwork}) = fieldtype(TN, :data_graph)
-DataGraphs.underlying_graph(tn::SimpleTensorNetwork) = underlying_graph(data_graph(tn))
-DataGraphs.underlying_graph_type(TN::Type{<:SimpleTensorNetwork}) =
+data_graph(tn::TensorNetwork) = getfield(tn, :data_graph)
+data_graph_type(TN::Type{<:TensorNetwork}) = fieldtype(TN, :data_graph)
+DataGraphs.underlying_graph(tn::TensorNetwork) = underlying_graph(data_graph(tn))
+DataGraphs.underlying_graph_type(TN::Type{<:TensorNetwork}) =
     fieldtype(data_graph_type(TN), :underlying_graph)
-DataGraphs.vertex_data(graph::SimpleTensorNetwork, args...) =
+DataGraphs.vertex_data(graph::TensorNetwork, args...) =
     vertex_data(data_graph(graph), args...)
-DataGraphs.edge_data(graph::SimpleTensorNetwork, args...) =
-    edge_data(data_graph(graph), args...)
+DataGraphs.edge_data(graph::TensorNetwork, args...) = edge_data(data_graph(graph), args...)
 
-function Base.setindex!(tn::SimpleTensorNetwork, t::AbstractIndexedArray, v::Int)
+function Base.setindex!(tn::TensorNetwork, t::AbstractIndexedArray, v::Int)
     tn.data_graph[v] = t
 end
 
-Base.getindex(tn::SimpleTensorNetwork, v::Int) = tn.data_graph[v]
+Base.getindex(tn::TensorNetwork, v::Int) = tn.data_graph[v]
 
 """
 Return if a tensor network `tn` has a cycle. If it has not a cycle, `tn` is a tree tensor network.
 """
-Graphs.is_cyclic(tn::SimpleTensorNetwork) =
+Graphs.is_cyclic(tn::TensorNetwork) =
     Graphs.is_cyclic(tn.data_graph.underlying_graph.position_graph)
 
 
-Graphs.has_edge(tn::SimpleTensorNetwork, e::NamedEdge) = Graphs.has_edge(tn.data_graph, e)
+Graphs.has_edge(tn::TensorNetwork, e::NamedEdge) = Graphs.has_edge(tn.data_graph, e)
 
 """
 Contract all the tensors in a tensor network `tn` and return the result.
@@ -68,7 +66,7 @@ This function works only for tree tensor networks, i.e., `is_cyclic(tn) == false
 
 root_vertex: The vertex to start the contraction. The default is 1.
 """
-function complete_contraction(tn::SimpleTensorNetwork; root_vertex::Int = 1)
+function complete_contraction(tn::TensorNetwork; root_vertex::Int = 1)
     !Graphs.is_cyclic(tn) ||
         error("complete_contraction is not supported only for a tree tensor network.")
     res = tn[root_vertex]
@@ -84,7 +82,7 @@ Contract all the tensors in a subtree of a tensor network `tn` and return the re
 The subtree is defined by a vertex `v` and its parent vertex `parent_v`.
 Note that `parent_v` is not included in the subtree.
 """
-function _contract_subtree(tn::SimpleTensorNetwork, v::Int, parent_v::Union{Int,Nothing})
+function _contract_subtree(tn::TensorNetwork, v::Int, parent_v::Union{Int,Nothing})
     res = tn[v]
     for nv in neighbors(tn.data_graph, v)
         if nv != parent_v

test/simpletensornetwork_test.jl

@@ -1,6 +1,5 @@
 @testitem "simpletensornetwork.jl" begin
-    import SimpleTensorNetworks:
-        Index, dim, IndexedArray, indices, permute, SimpleTensorNetwork
+    import SimpleTensorNetworks: Index, dim, IndexedArray, indices, permute, TensorNetwork
     import Graphs: is_connected, has_edge
 
     @testset "Construction from IndexedArray objects" begin
@@ -14,7 +13,7 @@
         t3 = IndexedArray(rand(2, 2), [c, d])
 
 
-        tn = SimpleTensorNetwork([t1, t2, t3])
+        tn = TensorNetwork([t1, t2, t3])
 
         @test has_edge(tn, 1 => 2)
         @test has_edge(tn, 2 => 1)
@@ -34,7 +33,7 @@
         t3 = IndexedArray(rand(2), [b])
         t4 = IndexedArray(rand(2), [c])
 
-        tn = SimpleTensorNetwork([t1, t2, t3, t4])
+        tn = TensorNetwork([t1, t2, t3, t4])
         @test only(SimpleTensorNetworks.complete_contraction(tn; root_vertex = 1)) ≈
               only(t1 * t2 * t3 * t4)
     end