better organize src files

Author: dehann

src/IncrementalInference.jl

@@ -60,7 +60,7 @@ using Logging
 using PrecompileTools
 
 # bringing in BSD 3-clause ccolamd
-include("ccolamd.jl")
+include("services/ccolamd.jl")
 using SuiteSparse.CHOLMOD: SuiteSparse_long # For CCOLAMD constraints.
 using .Ccolamd
 
@@ -120,8 +120,8 @@ include("ExportAPI.jl")
 # FIXME, move up to DFG
 # abstract type AbstractManifoldMinimize <: AbstractRelative end
 
-#
-include("ManifoldsExtentions.jl")
+
+
 # regular
 include("entities/SolverParams.jl")
 
@@ -133,10 +133,15 @@ include("Factors/GenericMarginal.jl")
 include("entities/AliasScalarSampling.jl")
 include("entities/OptionalDensities.jl")
 include("entities/BeliefTypes.jl")
+
+#
+include("manifolds/services/ManifoldsExtentions.jl")
+include("manifolds/services/ManifoldSampling.jl")
+
 include("entities/FactorGradients.jl")
 
 # Statistics helpers on manifolds
-include("VariableStatistics.jl")
+include("services/VariableStatistics.jl")
 
 # factors needed for belief propagation on the tree
 include("Factors/MsgPrior.jl")
@@ -201,44 +206,44 @@ include("services/ExplicitDiscreteMarginalizations.jl")
 include("services/EvalFactor.jl")
 include("services/ApproxConv.jl")
 
-# FIXME CONSOLIDATE
-include("ConsolidateParametricRelatives.jl")
 
 include("services/GraphProductOperations.jl")
 include("services/SolveTree.jl")
 include("services/TetherUtils.jl")
 include("services/TreeDebugTools.jl")
 include("CliqueStateMachine/services/CliqStateMachineUtils.jl")
 
+# FIXME CONSOLIDATE
+include("parametric/services/ConsolidateParametricRelatives.jl")
 #EXPERIMENTAL parametric
-include("ParametricCSMFunctions.jl")
-include("ParametricUtils.jl")
-include("ParametricManoptDev.jl")
+include("parametric/services/ParametricCSMFunctions.jl")
+include("parametric/services/ParametricUtils.jl")
+include("parametric/services/ParametricOptim.jl")
+include("parametric/services/ParametricManoptDev.jl")
 include("services/MaxMixture.jl")
 
 #X-stroke
 include("CliqueStateMachine/services/CliqueStateMachine.jl")
 
-include("CanonicalGraphExamples.jl")
+include("services/CanonicalGraphExamples.jl")
 
 include("services/AdditionalUtils.jl")
-include("SolverAPI.jl")
+include("services/SolverAPI.jl")
 
 # Symbolic tree analysis files.
-include("AnalysisTools.jl")
+include("services/AnalysisTools.jl")
 
-include("ManifoldSampling.jl")
 
 # deprecation legacy support
 include("Deprecated.jl")
 
 exportimg(pl) = error("Please do `using Gadfly` to allow image export.")
 function __init__()
   @require InteractiveUtils = "b77e0a4c-d291-57a0-90e8-8db25a27a240" include(
-    "RequireInteractiveUtils.jl",
+    "services/RequireInteractiveUtils.jl",
   )
   @require Gadfly = "c91e804a-d5a3-530f-b6f0-dfbca275c004" include(
-    "EmbeddedPlottingUtils.jl",
+    "services/EmbeddedPlottingUtils.jl",
   )
   @require DifferentialEquations = "0c46a032-eb83-5123-abaf-570d42b7fbaa" include(
     "ODE/DERelative.jl",

src/ManifoldSampling.jl renamed to src/manifolds/services/ManifoldSampling.jl

@@ -23,27 +23,6 @@ function Optim.project_tangent!(M::ManifoldWrapper, g, x)
     return g
 end
 
-# experimental
-function optimizeManifold_FD(
-  M::AbstractManifold, 
-  cost::Function,
-  x0::AbstractArray;
-  algorithm = Optim.ConjugateGradient(; manifold=ManifoldWrapper(M))
-)
-  # finitediff setup
-  r_backend = ManifoldDiff.TangentDiffBackend(
-    ManifoldDiff.FiniteDifferencesBackend()
-  )
-  
-  ## finitediff gradient (non-manual)
-  function costgrad_FD!(X,p)
-    X .= ManifoldDiff.gradient(M, cost, p, r_backend)
-    X
-  end
-
-  Optim.optimize(cost, costgrad_FD!, x0, algorithm)
-end
-
 
 ## ================================================================================================
 ## AbstractPowerManifold with N as field to avoid excessive compiling time.

src/ManifoldsExtentions.jl renamed to src/manifolds/services/ManifoldsExtentions.jl

@@ -0,0 +1,21 @@
+
+# experimental
+function optimizeManifold_FD(
+  M::AbstractManifold, 
+  cost::Function,
+  x0::AbstractArray;
+  algorithm = Optim.ConjugateGradient(; manifold=ManifoldWrapper(M))
+)
+  # finitediff setup
+  r_backend = ManifoldDiff.TangentDiffBackend(
+    ManifoldDiff.FiniteDifferencesBackend()
+  )
+  
+  ## finitediff gradient (non-manual)
+  function costgrad_FD!(X,p)
+    X .= ManifoldDiff.gradient(M, cost, p, r_backend)
+    X
+  end
+
+  Optim.optimize(cost, costgrad_FD!, x0, algorithm)
+end