Finish new TE and bump version (#446)

* improve docstring of transfer operator

* apply Runic

* bump version

* add missing docrefs to transfer operator

Author: Datseris

CHANGELOG.md

@@ -2,6 +2,10 @@
 
 Changelog is kept with respect to version 0.11 of Entropies.jl. From version v2.0 onwards, this package has been renamed to ComplexityMeasures.jl.
 
+## 3.10
+
+- `TransferOperator` has been reworked and substantially enhanced. It is now a `ProbabilitiesEstimator` and works with any count based outcome space. See its new docstring for more!
+
 ## 3.9
 
 - Adding a new feature to `PowerSpectrum` estimator that adds a threshold for "small" amplitudes. The threshold is applied to the probabilities by default with an option to apply it to the power spectrum.

Project.toml

@@ -2,7 +2,7 @@ name = "ComplexityMeasures"
 uuid = "ab4b797d-85ee-42ba-b621-05d793b346a2"
 authors = "Kristian Agasøster Haaga <kahaaga@gmail.com>, George Datseries <datseris.george@gmail.com>"
 repo = "https://github.com/juliadynamics/ComplexityMeasures.jl.git"
-version = "3.9.0"
+version = "3.10.0"
 
 [deps]
 Combinatorics = "861a8166-3701-5b0c-9a16-15d98fcdc6aa"

docs/make.jl

@@ -38,9 +38,10 @@ using DocumenterCitations
 
 bib = CitationBibliography(
     joinpath(@__DIR__, "refs.bib");
-    style=:authoryear
+    style = :authoryear
 )
 
-build_docs_with_style(pages, ComplexityMeasures, StateSpaceSets;
-    expandfirst = ["index.md"], bib,linkcheck=false
+build_docs_with_style(
+    pages, ComplexityMeasures, StateSpaceSets;
+    expandfirst = ["index.md"], bib, linkcheck = false
 )

docs/src/index.md

@@ -6,20 +6,12 @@ ComplexityMeasures
 
 ## Latest news
 
-ComplexityMeasures.jl has been updated to v3!
+[`TransferOperator`](@ref) has been reworked and substantially enhanced. It is now a `ProbabilitiesEstimator` and works with any count based outcome space. See its new docstring for more!
 
-The software has been massively improved and its core principles were
-redesigned to be extendable, accessible, and more closely based
-on the rigorous mathematics of probabilities and entropies.
+You can always check the
+[CHANGELOG.md](https://github.com/JuliaDynamics/ComplexityMeasures.jl/main/CHANGELOG.md) for details!
 
-For more details of this new release, please see our [announcement
-post on discourse](https://discourse.julialang.org/t/complexitymeasures-jl-v3-a-mathematically-rigorous-software-for-probability-entropy-and-complexity/108562)
-or the central [Tutorial](@ref) of the v3 documentation.
-
-In this v3 many concepts were renamed, but there is no formally
-breaking change. Everything that changed has been deprecated
-and is backwards compatible. You can see the
-[CHANGELOG.md](https://github.com/JuliaDynamics/ComplexityMeasures.jl/blob/v3.0.0/CHANGELOG.md) for more details!
+Note that the **DynamicalSystem.jl** library publishes general updates in a [JuliaLang Discourse post](https://discourse.julialang.org/t/dynamicalsystems-jl-news-updates-and-announcements/122079) that you can subscribe to!
 
 ## Documentation contents
 

docs/src/measure_count.jl

@@ -17,7 +17,7 @@ function concrete_subtypes!(out, type::Type)
     else
         foreach(T -> concrete_subtypes!(out, T), subtypes(type))
     end
-    out
+    return out
 end
 
 # ## Count Based Outcome Spaces
@@ -102,7 +102,7 @@ INFO_MEASURE_ESTIMATOR_GENERIC = concrete_subtypes(ComplexityMeasures.DiscreteIn
 
 # so we count by multiplying
 
-n_discrete_infoest_generic = length(INFO_MEASURES_DISCRETE)*length(INFO_MEASURE_ESTIMATOR_GENERIC)
+n_discrete_infoest_generic = length(INFO_MEASURES_DISCRETE) * length(INFO_MEASURE_ESTIMATOR_GENERIC)
 
 
 # In addition to the generic estimators,
@@ -196,7 +196,7 @@ n_extra_prob_measures = 2 * n_probs_discrete
 # ComplexityMeasures.jl are:
 
 n_grand_total =
-  n_discrete_info +
-  n_diff_info +
-  n_complexity_measures_total +
-  n_extra_prob_measures
+    n_discrete_info +
+    n_diff_info +
+    n_complexity_measures_total +
+    n_extra_prob_measures

docs/src/probabilities.md

@@ -121,7 +121,15 @@ RelativeAmount
 BayesianRegularization
 Shrinkage
 AddConstant
+```
+
+### Transfer operator
+
+```@docs
 TransferOperator
+transferoperator
+ApproximationIterative
+ApproximationEigen
 ```
 
 ## [Encodings/Symbolizations API](@id encodings)

docs/src/tutorial.jl

@@ -117,7 +117,7 @@ barplot(left_edges, probs; axis = (ylabel = "probability",))
 
 using DynamicalSystemsBase
 
-logistic_rule(u, r, t) = SVector(r*u[1]*(1 - u[1]))
+logistic_rule(u, r, t) = SVector(r * u[1] * (1 - u[1]))
 ds = DeterministicIteratedMap(logistic_rule, [0.4], 4.0)
 Y, t = trajectory(ds, 10_000; Ttr = 100)
 y = Y[:, 1]

src/ComplexityMeasures.jl

@@ -2,9 +2,9 @@ module ComplexityMeasures
 
 # Use the README as the module docs
 @doc let
-  path = joinpath(dirname(@__DIR__), "README.md")
-  include_dependency(path)
-  read(path, String)
+    path = joinpath(dirname(@__DIR__), "README.md")
+    include_dependency(path)
+    read(path, String)
 end ComplexityMeasures
 
 using Reexport

src/complexity_measures/approximate_entropy.jl

@@ -76,8 +76,10 @@ Base.@kwdef struct ApproximateEntropy{I, B, R} <: ComplexityEstimator
         r > 0 || throw(ArgumentError("r must be > 0. Got r=$(r)."))
         new{I, B, R}(m, τ, base, r)
     end
-    function ApproximateEntropy(x::AbstractVector; m::Int = 2, τ::Int = 1,
-            base = MathConstants.e, r = 0.2 * Statistics.std(x))
+    function ApproximateEntropy(
+            x::AbstractVector; m::Int = 2, τ::Int = 1,
+            base = MathConstants.e, r = 0.2 * Statistics.std(x)
+        )
         ApproximateEntropy(m, τ, base, r)
     end
 end
@@ -86,12 +88,12 @@ function complexity(c::ApproximateEntropy, x::AbstractStateSpaceSet)
     throw(ArgumentError("Approximate entropy is only computable for timeseries."))
 end
 
-function complexity(c::ApproximateEntropy, x::AbstractVector{T}) where T
+function complexity(c::ApproximateEntropy, x::AbstractVector{T}) where {T}
     (; m, τ, r, base) = c
 
     # Definition in https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7515030/
     if m == 1
-       return compute_ϕ(x; k = m, r, τ, base)
+        return compute_ϕ(x; k = m, r, τ, base)
     else
         ϕᵐ = compute_ϕ(x; k = m, r, τ, base)
         ϕᵐ⁺¹ = compute_ϕ(x; k = m + 1, r, τ, base)
@@ -125,14 +127,16 @@ C_i^k(r) = \\textrm{number of } j \\textrm{ such that } d({\\bf u}_i, {\\bf u}_j
 where ``d`` is the maximum (Chebyshev) distance, `r` is the tolerance, and `N` is the
 length of the original scalar-valued time series `x`.
 """
-function compute_ϕ(x::AbstractVector{T}; r = 0.2 * Statistics.std(x), k::Int = 2,
-        τ::Int = 1, base = MathConstants.e) where T <: Real
-    τs = 0:τ:(k - 1)*τ
+function compute_ϕ(
+        x::AbstractVector{T}; r = 0.2 * Statistics.std(x), k::Int = 2,
+        τ::Int = 1, base = MathConstants.e
+    ) where {T <: Real}
+    τs = 0:τ:((k - 1) * τ)
     pts = genembed(x, τs)
     tree = KDTree(pts, Chebyshev())
 
     # Account for `τ != 1` in the normalization constant.
-    f = length(x) - (k - 1)*τ
+    f = length(x) - (k - 1) * τ
 
     # `inrangecount` counts the query point itself, which is wanted for approximate entropy,
     # because there are always neighbors and thus log(0) is never encountered.

src/complexity_measures/bubble_entropy.jl

@@ -62,14 +62,14 @@ end
 function complexity_normalized(est::BubbleEntropy, x)
     o_m = BubbleSortSwaps(m = est.m)
     o_m₊₁ = BubbleSortSwaps(m = est.m + 1)
-    h_m =  information(est.definition, o_m, x)
-    h_m₊₁ =  information(est.definition, o_m₊₁, x)
+    h_m = information(est.definition, o_m, x)
+    h_m₊₁ = information(est.definition, o_m₊₁, x)
 
     # The normalized factor as (I think) described in Manis et al. (2017).
     # Their description is a bit unclear to me.
     h_max_m = information_maximum(est.definition, o_m, x)
     h_max_m₊₁ = information_maximum(est.definition, o_m₊₁, x)
     norm_factor = (h_max_m₊₁ - h_max_m) # maximum difference for dims `m` and `m + 1`
 
-    return (h_m₊₁ - h_m)/norm_factor
-end
+    return (h_m₊₁ - h_m) / norm_factor
+end