SDeMo QoL improvements (#289)

* feat(demo)?: maxdepth

* feat: respects depth/length when training

* doc: max

* feat(demo): feature bagging

* feat: export bag features

* bug: correct number of features

SDeMo/Project.toml

@@ -1,7 +1,7 @@
 name = "SDeMo"
 uuid = "3e5feb82-bcca-434d-9cd5-c11731a21467"
 authors = ["Timothée Poisot <timothee.poisot@umontreal.ca>"]
-version = "0.0.3"
+version = "0.0.4"
 
 [deps]
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"

SDeMo/src/SDeMo.jl

@@ -42,10 +42,11 @@ export BIOCLIM
 # BIOCLIM
 include("classifiers/decisiontree.jl")
 export DecisionTree
+export maxnodes!, maxdepth!
 
 # Bagging and ensembles
 include("ensembles/bagging.jl")
-export Bagging, outofbag, bootstrap
+export Bagging, outofbag, bootstrap, bagfeatures!
 
 include("ensembles/ensemble.jl")
 export Ensemble

SDeMo/src/classifiers/decisiontree.jl

@@ -76,18 +76,43 @@ end
 """
     DecisionTree
 
-TODO
+The depth and number of nodes can be adjusted with `maxnodes!` and `maxdepth!`.
 """
 Base.@kwdef mutable struct DecisionTree <: Classifier
     root::DecisionNode = DecisionNode()
     maxnodes::Integer = 12
+    maxdepth::Integer = 7
+end
+
+function maxnodes!(dt::DecisionTree, n::Integer)
+    dt.maxnodes = n
+    return dt
+end
+function maxdepth!(dt::DecisionTree, n::Integer)
+    dt.maxdepth = n
+    return dt
+end
+
+function maxnodes!(sdm::SDM, n)
+    if sdm.classifier isa DecisionTree
+        maxnodes!(sdm.classifier, n)
+        return sdm
+    end
+    return sdm
+end
+function maxdepth!(sdm::SDM, n)
+    if sdm.classifier isa DecisionTree
+        maxdepth!(sdm.classifier, n)
+        return sdm
+    end
+    return sdm
 end
 
 tips(::Nothing) = nothing
 tips(dt::DecisionTree) = tips(dt.root)
 function tips(dn::SDeMo.DecisionNode)
     if iszero(dn.variable)
-        return dn
+        return [dn]
     else
         return vcat(tips(dn.left), tips(dn.right))
     end
@@ -157,7 +182,7 @@ function train!(
     root.prediction = mean(y)
     dt.root = root
     train!(dt.root, X, y)
-    for _ in 1:6
+    for _ in 1:(dt.maxdepth-2)
         for tip in SDeMo.tips(dt)
             p = SDeMo._pool(tip, X)
             if !(tip.visited)
@@ -191,4 +216,14 @@ end
 
 function StatsAPI.predict(dt::DecisionTree, X::Matrix{T}) where {T <: Number}
     return vec(mapslices(x -> predict(dt, x), X; dims = 1))
+end
+
+@testitem "We can train a decison tree" begin
+    X, y = SDeMo.__demodata()
+    model = SDM(MultivariateTransform{PCA}, DecisionTree, X, y)
+    maxdepth!(model, 3)
+    @test model.classifier.maxdepth == 3
+    train!(model)
+    @test SDeMo.depth(model.classifier) <= 3
+    @test length(SDeMo.tips(model.classifier)) <= model.classifier.maxnodes
 end

SDeMo/src/ensembles/bagging.jl

@@ -1,11 +1,11 @@
 """
     bootstrap(y, X; n = 50)
 """
-function bootstrap(y, X; n=50)
+function bootstrap(y, X; n = 50)
     @assert size(y, 1) == size(X, 2)
     bags = []
     for _ in 1:n
-        inbag = sample(1:size(X, 2), size(X, 2); replace=true)
+        inbag = sample(1:size(X, 2), size(X, 2); replace = true)
         outbag = setdiff(axes(X, 2), inbag)
         push!(bags, (inbag, outbag))
     end
@@ -24,7 +24,7 @@ end
 """
 mutable struct Bagging <: AbstractEnsembleSDM
     model::SDM
-    bags::Vector{Tuple{Vector{Int64},Vector{Int64}}}
+    bags::Vector{Tuple{Vector{Int64}, Vector{Int64}}}
     models::Vector{SDM}
 end
 
@@ -43,7 +43,7 @@ end
 Creates a bag from SDM
 """
 function Bagging(model::SDM, n::Integer)
-    bags = bootstrap(labels(model), features(model); n=n)
+    bags = bootstrap(labels(model), features(model); n = n)
     return Bagging(model, bags, [deepcopy(model) for _ in eachindex(bags)])
 end
 
@@ -77,7 +77,27 @@ function outofbag(ensemble::Bagging; kwargs...)
     return ConfusionMatrix(outcomes, ensemble.model.y[done_instances])
 end
 
+bagfeatures!(ensemble::Bagging) =
+    bagfeatures!(ensemble, ceil(Int64, sqrt(length(variables(ensemble)))))
+
+function bagfeatures!(ensemble::Bagging, n::Integer)
+    for model in ensemble.models
+        sampled_variables = StatsBase.sample(variables(model), n; replace = false)
+        variables!(model, sampled_variables)
+    end
+    return ensemble
+end
+
+@testitem "We can bag the features of an ensemble model" begin
+    X, y = SDeMo.__demodata()
+    model = SDM(MultivariateTransform{PCA}, DecisionTree, X, y)
+    ensemble = Bagging(model, 10)
+    bagfeatures!(ensemble)
+    for model in ensemble.models
+        @test length(variables(model)) == ceil(Int64, sqrt(size(X, 1)))
+    end
+end
 
 majority(pred::Vector{Bool}) = sum(pred) > length(pred) // 2
 majority(pred::BitVector) = sum(pred) > length(pred) // 2
-export majority
+export majority

SDeMo/src/ensembles/ensemble.jl

@@ -41,4 +41,4 @@ Base.deleteat!(ens::Ensemble, i) = deleteat!(ens.models, i)
     outmod = popat!(ens, 1)
     @test outmod isa SDM
     @test length(ens.models) == 1
-end
+end

docs/src/tutorials/sdemo.jl

@@ -85,7 +85,7 @@ hm = heatmap!(ax,
 scatter!(ax, presencelayer; color = :black)
 scatter!(ax, bgpoints; color = :red, markersize = 4)
 lines!(ax, CHE.geometry[1]; color = :black)
-Colorbar(f[1,2], hm)
+Colorbar(f[1, 2], hm)
 hidedecorations!(ax)
 hidespines!(ax)
 current_figure() #hide
@@ -144,9 +144,11 @@ ensemble = Bagging(sdm, 30)
 # of the dataset, we can also bootstrap which variables are accessible to each
 # model:
 
-for model in ensemble.models
-    variables!(model, unique(rand(variables(model), length(variables(model)))))
-end
+bagfeatures!(ensemble)
+
+# By default, the `bagfeatures!` function called on an ensemble will sample the variables
+# forom the model, so that each model in the ensemble has the square root (rounded _up_) of
+# the number of original variables.
 
 # ::: info About this ensemble model
 # 
@@ -182,15 +184,25 @@ f = Figure(; size = (600, 600))
 ax = Axis(f[1, 1]; aspect = DataAspect(), title = "Prediction")
 hm = heatmap!(ax, prd; colormap = :linear_worb_100_25_c53_n256, colorrange = (0, 1))
 Colorbar(f[1, 2], hm)
-contour!(ax, predict(ensemble, layers; consensus=majority); color = :black, linewidth = 0.5)
+contour!(
+    ax,
+    predict(ensemble, layers; consensus = majority);
+    color = :black,
+    linewidth = 0.5,
+)
 lines!(ax, CHE.geometry[1]; color = :black)
 hidedecorations!(ax)
 hidespines!(ax)
 ax2 = Axis(f[2, 1]; aspect = DataAspect(), title = "Uncertainty")
 hm =
     heatmap!(ax2, quantize(unc); colormap = :linear_gow_60_85_c27_n256, colorrange = (0, 1))
 Colorbar(f[2, 2], hm)
-contour!(ax2, predict(ensemble, layers; consensus=majority); color = :black, linewidth = 0.5)
+contour!(
+    ax2,
+    predict(ensemble, layers; consensus = majority);
+    color = :black,
+    linewidth = 0.5,
+)
 lines!(ax2, CHE.geometry[1]; color = :black)
 hidedecorations!(ax2)
 hidespines!(ax2)
@@ -217,14 +229,24 @@ hm = heatmap!(
     colormap = :diverging_gwv_55_95_c39_n256,
     colorrange = (-0.3, 0.3),
 )
-contour!(ax, predict(ensemble, layers; consensus=majority); color = :black, linewidth = 0.5)
+contour!(
+    ax,
+    predict(ensemble, layers; consensus = majority);
+    color = :black,
+    linewidth = 0.5,
+)
 lines!(ax, CHE.geometry[1]; color = :black) #hide
 hidedecorations!(ax)
 hidespines!(ax)
 Colorbar(f[1, 2], hm)
 ax2 = Axis(f[2, 1]; aspect = DataAspect(), title = "Partial response")
 hm = heatmap!(ax2, part_v1; colormap = :linear_gow_65_90_c35_n256, colorrange = (0, 1))
-contour!(ax2, predict(ensemble, layers; consensus=majority); color = :black, linewidth = 0.5)
+contour!(
+    ax2,
+    predict(ensemble, layers; consensus = majority);
+    color = :black,
+    linewidth = 0.5,
+)
 lines!(ax2, CHE.geometry[1]; color = :black)
 Colorbar(f[2, 2], hm)
 hidedecorations!(ax2)
@@ -251,8 +273,13 @@ heatmap!(
         categorical = true,
     ),
 )
-contour!(ax, predict(ensemble, layers; consensus=majority); color = :black, linewidth = 0.5)
+contour!(
+    ax,
+    predict(ensemble, layers; consensus = majority);
+    color = :black,
+    linewidth = 0.5,
+)
 lines!(ax, CHE.geometry[1]; color = :black)
 hidedecorations!(ax)
 hidespines!(ax)
-current_figure() #hide
+current_figure() #hide