Improve performance for classification methods

examples/CO24/functions_classification.jl

@@ -215,7 +215,7 @@ function misclassificationrate(
         threshold = mean(σ)
         classif = σ .>= threshold
     else
-        distances = [abs(σ[i] - σ[j]) for i in eachindex(σ), j in eachindex(σ)]
+        distances = MeanFieldGraph.pairwise_abs_distances(σ)
         ct = cutree(hclust(distances; linkage=clustering); k=2)
         id_excitatory = ct[argmax(σ)]
         classif = ct .== id_excitatory

src/MeanFieldGraph.jl

@@ -4,7 +4,7 @@ using Clustering: Clustering, ClusteringResult, assignments, counts, kmeans, hcl
 using Distributions: Distributions, Bernoulli, DiscreteUniform, fit, mean
 using LinearAlgebra: LinearAlgebra, I, transpose
 using Plots: Plots, heatmap, palette, plot
-using KrylovKit: KrylovKit, eigsolve
+using KrylovKit: KrylovKit, svdsolve
 
 export MarkovChainModel, DiscreteTimeData, MarkovChainConnectivity
 export mvw, mvw_inf

src/classification.jl

@@ -15,10 +15,10 @@ function classification(
     if method == :aggregated
         σ̂ = covariance_vector(data)
     elseif method == :spectral
-        # compute the leading singular vector of the covariance matrix
+        # compute the leading right singular vector of the covariance matrix
         Σ̂ = covariance_matrix(data)
-        _, vecs = eigsolve(transpose(Σ̂) * Σ̂) # faster than full SVD
-        v̌ = vecs[1]
+        _, _, rvecs, = svdsolve(Σ̂, 1) # avoids forming Σ̂'Σ̂ explicitly
+        v̌ = rvecs[1]
 
         # sign disambiguation
         σ̂_ag = sum(Σ̂; dims=1)[1, :]
@@ -40,7 +40,7 @@ function classification(
         threshold = mean(σ̂)
         output = σ̂ .>= threshold
     elseif clustering in (:single, :average, :complete, :ward)
-        distances = [abs(σ̂[i] - σ̂[j]) for i in eachindex(σ̂), j in eachindex(σ̂)]
+        distances = pairwise_abs_distances(σ̂)
         ct = cutree(hclust(distances; linkage=clustering); k=2)
         id_excitatory = ct[argmax(σ̂)]
         output = ct .== id_excitatory
@@ -69,17 +69,28 @@ function covariance_matrix(data::DiscreteTimeData)::Matrix{Float64}
     N, T = size(data)
     Z = sum(X; dims=2)
 
-    s = zeros((N, N))
-    for t in 1:(T - 1)
-        s += @views(X[:, t + 1] * transpose(X[:, t]))
-    end
+    s = @views(X[:, 2:end] * transpose(X[:, 1:(end - 1)]))
     output = s / (T - 1) - Z * transpose(Z) / T^2
 
     return output
 end
 
 # Auxiliary functions
 
+function pairwise_abs_distances(v::Vector{<:Real})::Matrix{Float64}
+    n = length(v)
+    output = Matrix{Float64}(undef, n, n)
+    @inbounds for i in 1:n
+        output[i, i] = 0.0
+        @inbounds for j in (i + 1):n
+            d = abs(v[i] - v[j])
+            output[i, j] = d
+            output[j, i] = d
+        end
+    end
+    return output
+end
+
 function cluster2bool(R::ClusteringResult)::Vector{Bool}
     output = Vector{Bool}(undef, sum(counts(R)))
     check = R.centers[1] < R.centers[2]