Testing (#17)

* testing full graph and multi spin nodes

* new testing

* testing on larger data sets given in fils

* done tests on grid, draft of tests on chimera

* correct some testing

* some refactoring of tensor generation and some tests

* better dirac delta implementation in tensor forming

* first tests on chimera

* additional testings on chimera

* testing on chimera and raport

* some testing on chimers

* add testing file

* some notes concerning refactoring

* a correction

* tests on chimera 2048

* started refactoring compute_sungle_tensor()

* correction

* incorporating energy into the grid

* refactored mps construction

* refactor notation, prepare for grid implemetation, refactor tensor creation for peps

* correction, that should speed up

* an approach with interaction matrices

* first approach to graph representation

* some developments of graph notation

* refactored compute_single_tensor() a high speedup of computationn

* advance in coding graph

* first implementation of peps on graph

* correction

* peps calculated from a graph

* correction

* refctor notations

* changed and tested compression to these from SpinGlass

* corrections

* permute indexes of tensor in peps formation

* cleare peps_no_type

* correction

* correction

* correction

* some moer refactioring of peps

* change probability computation in peps

* introduce MPS MPO types to mpo-mps

* correction

* partial refactoring of mpo-mps

* new brute force in tests

* isne ising_graph and some new brute force aplications

* remove redundant functions, use energy and brute_force

* correct types

* coding for largel chimera cells

* correct the multi nodes chimera

* correct tests

* use bruteforce for internal energy and simplify notations

* extend reindexinf functions for the low energy spectrum approximation

* a correction

* remove some of testing instances and binary files

* correction

* move notations

* move tests to right place

* remove redundant runtests.jl

* remove generic linear algebra

* merget with master, test does not pass as it was before

* remove includes

* correct imports

* correction

* correct and speed up tests

* removed duplicated test on instance

Co-authored-by: Łukasz Pawela <lukasz.pawela@gmail.com>

Project.toml

@@ -16,6 +16,11 @@ Requires = "ae029012-a4dd-5104-9daa-d747884805df"
 TensorCast = "02d47bb6-7ce6-556a-be16-bb1710789e2b"
 TensorOperations = "6aa20fa7-93e2-5fca-9bc0-fbd0db3c71a2"
 
+Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+Logging = "56ddb016-857b-54e1-b83d-db4d58db5568"
+Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
+NPZ =  "15e1cf62-19b3-5cfa-8e77-841668bca605"
+
 [compat]
 TensorOperations = "3.0.1"
 julia = "1.4, 1.5"

src/SpinGlassPEPS.jl

@@ -12,12 +12,15 @@ module SpinGlassPEPS
 
     include("base.jl")
     include("compressions.jl")
-    include("contractions.jl")   
+    include("contractions.jl")
     include("ising.jl")
     include("graph.jl")
     # include("PEPS.jl")
     include("search.jl")
     include("utils.jl")
+    include("notation.jl")
+    include("peps_no_types.jl")
+    include("mps_implementation.jl")
 
     function __init__()
         @require CUDA="052768ef-5323-5732-b1bb-66c8b64840ba" begin
@@ -29,10 +32,10 @@ module SpinGlassPEPS
                 const CuQR = CUDA.CUSOLVER.CuQR
                 # scalar indexing is fine before 0.2
                 # CUDA.allowscalar(false)
-                include("cuda/base.jl") 
+                include("cuda/base.jl")
                 include("cuda/contractions.jl")
                 include("cuda/compressions.jl")
             end
         end
     end
-end
+end