-
Notifications
You must be signed in to change notification settings - Fork 7
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
* fixed tests w.r.t. deprecations in Peaks * fixed samplesize-bug; new default for samplesize; adjusted tests * rm Peaks.jl-dep; revised tests for mdop and garcia
- Loading branch information
Showing
6 changed files
with
109 additions
and
128 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,103 +1,80 @@ | ||
| using DelayEmbeddings, DynamicalSystemsBase | ||
| using Test | ||
| using Random | ||
| import Peaks | ||
|
|
||
| @testset "Pecora" begin | ||
| # %% Generate data | ||
| lor = Systems.lorenz([0.0;1.0;0.0];ρ=60) | ||
| data = trajectory(lor, 200; Δt=0.02, Ttr = 10) | ||
| metric = Chebyshev() | ||
|
|
||
| UNDERSAMPLING = false | ||
|
|
||
| @testset "Pecora univariate" begin | ||
| # %% Timeseries case | ||
|
|
||
| s = data[:, 2] # input timeseries = first entry of lorenz | ||
| optimal_τ = estimate_delay(s, "mi_min") | ||
| Tmax = 100 | ||
| K = 14 | ||
| samplesize = 1 | ||
| s = data[:, 2] # input timeseries = first entry of lorenz | ||
| optimal_τ = estimate_delay(s, "mi_min") | ||
| Tmax = 100 | ||
| K = 14 | ||
| samplesize = 1 | ||
|
|
||
| # using PyPlot | ||
| # pygui(true) | ||
| # figure() | ||
| # ax1 = subplot(211) | ||
| # ylabel("⟨ε★⟩") | ||
| # ax2 = subplot(212) | ||
| # xlabel("τ (index units)") | ||
| # ylabel("⟨Γ⟩") | ||
| τs = (0,) | ||
| Random.seed!(123) | ||
| es_ref, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric) | ||
| (max1,_) = DelayEmbeddings.findlocalextrema(vec(es_ref)) | ||
| @test optimal_τ - 2 ≤ max1[1]-1 ≤ optimal_τ + 2 | ||
| maxi = maximum(es_ref) | ||
| @test maxi ≤ 1.3 | ||
|
|
||
| τs = (0,) | ||
| Random.seed!(123) | ||
| es_ref, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric, undersampling = UNDERSAMPLING) | ||
| (max1,_) = Peaks.findmaxima(vec(es_ref)) | ||
| @test optimal_τ - 2 ≤ max1[1]-1 ≤ optimal_τ + 2 | ||
| maxi = maximum(es_ref) | ||
| @test maxi ≤ 1.3 | ||
| # ax1.plot(es_ref, label = "τs = $(τs)") | ||
| # ax2.plot(Γs) | ||
| Random.seed!(123) | ||
| es_ref2, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, K = K, metric = metric) | ||
| (max2,_) = DelayEmbeddings.findlocalextrema(vec(es_ref2)) | ||
| @test optimal_τ - 2 ≤ max2[1]-1 ≤ optimal_τ + 2 | ||
| maxi2 = maximum(es_ref2) | ||
| @test maxi - .1 ≤ maxi2 ≤ maxi + .1 | ||
|
|
||
| τs = (0, max1[1]-1,) | ||
| Random.seed!(123) | ||
| es, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric, undersampling = UNDERSAMPLING) | ||
| (max2,_) = Peaks.findmaxima(vec(es)) | ||
| (min1,_) = Peaks.findminima(vec(es)) | ||
| @test 1 ≤ max2[1]-1 ≤ 4 | ||
| @test optimal_τ - 2 ≤ min1[1]-1 ≤ optimal_τ + 2 | ||
| # ax1.plot(es, label = "τs = $(τs)") | ||
| # ax2.plot(Γs) | ||
| τs = (0, max1[1]-1,) | ||
| Random.seed!(123) | ||
| es, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric) | ||
| (max2,min1) = DelayEmbeddings.findlocalextrema(vec(es)) | ||
| @test 1 ≤ max2[1]-1 ≤ 4 | ||
| @test optimal_τ - 2 ≤ min1[2]-1 ≤ optimal_τ + 2 | ||
|
|
||
| τs = (0, max1[1]-1, max2[1]-1) | ||
| Random.seed!(123) | ||
| es, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric, undersampling = UNDERSAMPLING) | ||
| (min2,_) = Peaks.findminima(vec(es)) | ||
| @test max2[1]-2 ≤ min2[1]-1 ≤ max2[1] | ||
| # ax1.plot(es, label = "τs = $(τs)") | ||
| # ax2.plot(Γs) | ||
| # ax1.legend() | ||
| # ax1.set_title("lorenz data") | ||
| τs = (0, max1[1]-1, max2[1]-1) | ||
| Random.seed!(123) | ||
| es, Γs = pecora(s, τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric) | ||
| (_,min2) = DelayEmbeddings.findlocalextrema(vec(es)) | ||
| @test max2[1]-2 ≤ min2[1]-1 ≤ max2[1] | ||
|
|
||
| end | ||
|
|
||
| @testset "Pecora multivariate" begin | ||
| ## %% Trajectory case | ||
| s = Dataset(data) | ||
| optimal_τ = estimate_delay(s[:,2], "mi_min") | ||
| Tmax = 100 | ||
| K = 14 | ||
| samplesize = 1 | ||
| s = Dataset(data) | ||
| optimal_τ = estimate_delay(s[:,2], "mi_min") | ||
| Tmax = 100 | ||
| K = 14 | ||
| samplesize = 1 | ||
|
|
||
| js = (2,) | ||
| τs = (0,) | ||
| Random.seed!(123) | ||
| es_ref, _ = pecora(s[:,2], τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric) | ||
| Random.seed!(123) | ||
| es_ref2, _ = pecora(s, τs, js; delays = 0:Tmax, w = optimal_τ, samplesize = .2, K = K, metric = metric) | ||
| Random.seed!(123) | ||
| es, _= pecora(s, τs, js; delays = 0:Tmax, samplesize = samplesize, w = optimal_τ, K = K, metric = metric) | ||
|
|
||
| js = (2,) | ||
| τs = (0,) | ||
| Random.seed!(123) | ||
| es_ref, Γs = pecora(s[:,2], τs; delays = 0:Tmax, w = optimal_τ, samplesize = samplesize, K = K, metric = metric, undersampling = UNDERSAMPLING) | ||
| Random.seed!(123) | ||
| es, Γs = pecora(s, τs, js; delays = 0:Tmax, samplesize = samplesize, w = optimal_τ, K = K, metric = metric, undersampling = UNDERSAMPLING) | ||
|
|
||
| @test round.(es[:,2], digits = 4) == round.(vec(es_ref),digits = 4) | ||
| (x_maxi,_) = Peaks.findmaxima(es[:,1]) | ||
| @test 8 ≤ x_maxi[1]-1 ≤ 10 | ||
| (z_maxi,_) = Peaks.findmaxima(es[:,3]) | ||
| @test 13 ≤ z_maxi[1]-1 ≤ 15 | ||
| @test round.(es[:,2], digits = 4) == round.(vec(es_ref),digits = 4) | ||
| (x_maxi,_) = DelayEmbeddings.findlocalextrema(es[:,1]) | ||
| @test 8 ≤ x_maxi[2]-1 ≤ 10 | ||
| (z_maxi,_) = DelayEmbeddings.findlocalextrema(es[:,3]) | ||
| @test 13 ≤ z_maxi[2]-1 ≤ 15 | ||
| (x_maxi_ref,_) = DelayEmbeddings.findlocalextrema(es[:,2]) | ||
| (x_maxi_ref2,_) = DelayEmbeddings.findlocalextrema(es_ref2[:,2]) | ||
|
|
||
| # using PyPlot | ||
| # pygui(true) | ||
| # figure() | ||
| # subplot(211) | ||
| # for i in 1:3 | ||
| # plot(es[:, i], label = "τs = $(τs), js = $(js), J = $(i)") | ||
| # end | ||
| # ylabel("⟨ε★⟩") | ||
| # title("lorenz system, multiple timeseries") | ||
| # legend() | ||
| # grid() | ||
| # subplot(212) | ||
| # plot(es_ref, label = "τs = $(τs), js = $(js), J = 2") | ||
| # ylabel("⟨ε★⟩-ref") | ||
| # xlabel("τ (index units)") | ||
| # grid() | ||
| @test x_maxi_ref[1:4] == x_maxi_ref2[1:4] | ||
| @test es[x_maxi_ref[1:4],2] .- .1 ≤ es_ref2[x_maxi_ref2[1:4],2] ≤ es[x_maxi_ref[1:4],2] .+ .1 | ||
|
|
||
| end | ||
| end |