Merge pull request #81 from JuliaDynamics/bbnue_filter_lags

Add possibility to exclude certain lags when selecting variables for `bbnue`

Project.toml

@@ -1,7 +1,7 @@
 name = "TransferEntropy"
 uuid = "ea221983-52f3-5440-99c7-13ea201cd633"
 repo = "https://github.com/kahaaga/TransferEntropy.jl.git"
-version = "1.4.0"
+version = "1.5.0"
 
 [deps]
 DSP = "717857b8-e6f2-59f4-9121-6e50c889abd2"

src/transferentropy/autoutils.jl

@@ -4,8 +4,9 @@ using DelayEmbeddings, Statistics
     construct_candidate_variables(
         source::Vector{AbstractVector}, 
         target::Vector{AbstractVector}, 
-        cond::Vector{AbstractVector};
+        [cond::Vector{AbstractVector}];
         k::Int = 1, include_instantaneous = true,
+        τexclude::Union{Int, Nothing} = nothing,
         maxlag::Union{Int, Float64} = 0.05
     ) → ([τs_source, τs_target, τs_cond, ks_targetfuture], [js_source, js_target, js_cond, js_targetfuture])
 
@@ -18,9 +19,13 @@ the variables.
 
 If `maxlag` is an integer, `maxlag` is taken as the maximum allowed embedding lag. If `maxlag` is a float, 
 then the maximum embedding lag is taken as `maximum([length.(source); length.(target); length.(cond)])*maxlag`.
+
+If `τexclude` is an integer, all variables whose embedding lag has absolute value equal to `exclude` will be 
+excluded.
 """
 function construct_candidate_variables(source, target, cond;
-        k::Int = 1, 
+        k::Int = 1,
+        τexclude::Union{Int, Nothing} = nothing,
         include_instantaneous = true,
         method_delay = "ac_min",
         maxlag::Union{Int, Float64} = 0.05)
@@ -39,9 +44,10 @@ function construct_candidate_variables(source, target, cond;
     τsmax_source = [estimate_delay(s, method_delay, τs) for s in source]
     τsmax_target = [estimate_delay(t, method_delay, τs) for t in target]
     τsmax_cond = [estimate_delay(c, method_delay, τs) for c in cond]
-     
+
     # Generate candidate set
     startlag = include_instantaneous ? 0 : -1
+
     τs_source = [[startlag:-1:-τ...,] for τ in τsmax_source]
     τs_target = [[startlag:-1:-τ...,] for τ in τsmax_target]
     τs_cond = [[startlag:-1:-τ...,] for τ in τsmax_cond]
@@ -50,15 +56,31 @@ function construct_candidate_variables(source, target, cond;
     js_targetfuture = [i for i in length(τs_source)+1:length(τs_source)+length(τs_target)]
     τs = [τs_source..., τs_target..., τs_cond...]
     js = [[i for x in 1:length(τs[i])] for i = 1:length(τs)]
-
+
+    # Variable filtering, if desired
+    if τexclude isa Int
+        τs = [filtered_τs(τsᵢ, jsᵢ, τexclude) for (τsᵢ, jsᵢ) in zip(τs, js)]
+        js = [filtered_js(τsᵢ, jsᵢ, τexclude) for (τsᵢ, jsᵢ) in zip(τs, js)]
+    end
     return [τs..., ks_targetfuture], [js..., js_targetfuture]
 end
 
+# Usaully, we use all lags from startlag:-\tau_max to construct variables. In some situations,
+# we may want to exclude som of those variables. 
+function filtered_τs(τs::AbstractVector{Int}, js::AbstractVector{Int}, τexclude::Int)
+    [τ for τ in τs if abs(τ) != abs.(τexclude)]
+end
+
+function filtered_js(τs::AbstractVector{Int}, js::AbstractVector{Int}, τexclude::Int)
+    [j for (τ, j) in zip(τs, js) if abs(τ) != abs.(τexclude)]
+end
+
 # source & target variant 
 function construct_candidate_variables(source, target; 
         k::Int = 1, 
+        τexclude::Union{Int, Nothing} = nothing,
         include_instantaneous = true,
-        method_delay = "mi_min",
+        method_delay = "ac_min",
         maxlag::Union{Int, Float64} = 0.05)
 
     # Ensure all time series are of the same length.
@@ -74,7 +96,7 @@ function construct_candidate_variables(source, target;
     # Find the maximum allowed embedding lag for each of the candidates.
     τsmax_source = [estimate_delay(s, method_delay, τs) for s in source]
     τsmax_target = [estimate_delay(t, method_delay, τs) for t in target]
-    
+
     # Generate candidate set
     startlag = include_instantaneous ? 0 : -1
     τs_source = [[startlag:-1:-τ...,] for τ in τsmax_source]
@@ -84,6 +106,12 @@ function construct_candidate_variables(source, target;
     js_targetfuture = [i for i in length(τs_source)+1:length(τs_source)+length(τs_target)]
     τs = [τs_source..., τs_target...,]
     js = [[i for x in 1:length(τs[i])] for i = 1:length(τs)]
+
+    # Variable filtering, if desired
+    if τexclude isa Int
+        τs = [filtered_τs(τsᵢ, jsᵢ, τexclude) for (τsᵢ, jsᵢ) in zip(τs, js)]
+        js = [filtered_js(τsᵢ, jsᵢ, τexclude) for (τsᵢ, jsᵢ) in zip(τs, js)]
+    end
 
     return [τs..., ks_targetfuture], [js..., js_targetfuture]
 end

test/runtests.jl

@@ -54,6 +54,7 @@ end
 
 @testset "Transfer entropy" begin 
     s, t, c = rand(100), rand(100), rand(100)
+
     println("Starting transfer entropy tests...")
     @testset "Generalized Renyi transfer entropy" begin
         # Straight-forward estimators
@@ -120,6 +121,32 @@ end
     end
 
     @testset "Automated estimators" begin
+        @testset "Variable exclusion" begin
+             # Use periodic signals, so we also can test variable selection methods,
+            # which for sensible testing, need to have their autocorrelation minima 
+            # > 1.
+            s = sin.(1:100) .+ rand(100)
+            t = sin.(1:100) .+ rand(100)
+            c = sin.(1:100) .+ rand(100)
+
+            τexclude = 1
+            vars = TransferEntropy.construct_candidate_variables([s], [t], τexclude = nothing)
+            vars_ex = TransferEntropy.construct_candidate_variables([s], [t], τexclude = τexclude)
+            fvars = Iterators.flatten(vars[1][1:end-1]) |> collect
+            fvars_ex = Iterators.flatten(vars_ex[1][1:end-1]) |> collect
+            @test τexclude ∈ abs.(fvars)
+            @test τexclude ∉ abs.(fvars_ex)
+            @test length(fvars) > length(fvars_ex)
+
+            vars = TransferEntropy.construct_candidate_variables([s], [t], [c], τexclude = nothing)
+            vars_ex = TransferEntropy.construct_candidate_variables([s], [t], [c], τexclude = τexclude)
+            fvars = Iterators.flatten(vars[1][1:end-1]) |> collect
+            fvars_ex = Iterators.flatten(vars_ex[1][1:end-1]) |> collect
+            @test τexclude ∈ abs.(fvars)
+            @test τexclude ∉ abs.(fvars_ex)
+            @test length(fvars) > length(fvars_ex)
+        end
+
         est = VisitationFrequency(RectangularBinning(3))
         te_st, params_st = bbnue(s, t, est)
         te_stc, params_stc = bbnue(s, t, c, est)