From 0161f888df2de8c877f906f01c9d9f54955c17d1 Mon Sep 17 00:00:00 2001
From: Or Dinari <dinari.or@gmail.com>
Date: Sun, 22 Aug 2021 20:04:40 +0300
Subject: [PATCH] fmeasure

---
 src/Clustering.jl |  6 +++-
 src/fmeasure.jl   | 83 +++++++++++++++++++++++++++++++++++++++++++++++
 test/fmeasure.jl  | 20 ++++++++++++
 test/runtests.jl  |  3 +-
 4 files changed, 110 insertions(+), 2 deletions(-)
 create mode 100644 src/fmeasure.jl
 create mode 100644 test/fmeasure.jl

diff --git a/src/Clustering.jl b/src/Clustering.jl
index 39e42783..a09d538e 100644
--- a/src/Clustering.jl
+++ b/src/Clustering.jl
@@ -64,7 +64,10 @@ module Clustering
     Hclust, hclust, cutree,
 
     # MCL
-    mcl, MCLResult
+    mcl, MCLResult,
+
+    #fmeasure
+    pair_precision, fmeasure, pair_recall
 
     ## source files
 
@@ -84,6 +87,7 @@ module Clustering
     include("varinfo.jl")
     include("vmeasure.jl")
     include("mutualinfo.jl")
+    include("fmeasure.jl")
 
     include("hclust.jl")
 
diff --git a/src/fmeasure.jl b/src/fmeasure.jl
new file mode 100644
index 00000000..9c0d4f0b
--- /dev/null
+++ b/src/fmeasure.jl
@@ -0,0 +1,83 @@
+"""
+    _pair_confusion_matrix(a,b) -> NTuple{4, Int64}
+
+Compute the similarities between two clusterings by considering all pairs of samples.
+
+Returns (TruePositive,FalsePositive,FalseNegative,TrueNegative)
+
+"""
+function _pair_confusion_matrix(a,b)
+    c = counts(a, b)
+    n = length(a)
+    n_k = sum(c,dims=1)[:]
+    n_c = sum(c,dims=2)[:]
+    n_sum = sum(c.*c)
+    tp = n_sum-n
+    fp = sum(c*n_k)-n_sum
+    fn = sum(c'*n_c)-n_sum
+    tn = n^2-fp-fn-n_sum 
+    return tp,fp,fn,tn
+end
+
+
+
+"""
+    pair_precision(a, b) -> Float64
+
+Compute the pair counting precision between two clustering of the same data points.
+
+`a` and `b` can be either [`ClusteringResult`](@ref) instances or
+assignments vectors (`AbstractVector{<:Integer}`).
+
+Returns the value of the pair counting precision.
+
+# References
+> Pfitzner, Darius, Richard Leibbrandt, and David Powers. (2009).
+> *Characterization and evaluation of similarity measures for pairs of clusterings.*
+> Knowledge and Information Systems: 361-394.
+"""
+function pair_precision(a, b)
+    tp,fp,fn,tn = _pair_confusion_matrix(a,b)
+    return tp/(tp+fp)
+end
+
+"""
+    pair_recall(a, b) -> Float64
+
+Compute the pair counting recall between two clustering of the same data points.
+
+`a` and `b` can be either [`ClusteringResult`](@ref) instances or
+assignments vectors (`AbstractVector{<:Integer}`).
+
+Returns the value of the pair counting recall.
+
+# References
+> Pfitzner, Darius, Richard Leibbrandt, and David Powers. (2009).
+> *Characterization and evaluation of similarity measures for pairs of clusterings.*
+> Knowledge and Information Systems: 361-394.
+"""
+function pair_recall(a, b)
+    tp,fp,fn,tn = _pair_confusion_matrix(a,b)
+    return tp/(tp+fn)
+end
+
+"""
+    fmeasure(a, b) -> Float64
+
+Compute the pair counting fmeasure between two clustering of the same data points.
+
+`a` and `b` can be either [`ClusteringResult`](@ref) instances or
+assignments vectors (`AbstractVector{<:Integer}`).
+
+Returns the value of the pair counting recall.
+
+# References
+> Pfitzner, Darius, Richard Leibbrandt, and David Powers. (2009).
+> *Characterization and evaluation of similarity measures for pairs of clusterings.*
+> Knowledge and Information Systems: 361-394.
+"""
+function fmeasure(a, b)
+    p = pair_precision(a,b)
+    r = pair_recall(a,b)
+    return (2*p*r)/(p+r)
+end
\ No newline at end of file
diff --git a/test/fmeasure.jl b/test/fmeasure.jl
new file mode 100644
index 00000000..2757261d
--- /dev/null
+++ b/test/fmeasure.jl
@@ -0,0 +1,20 @@
+using Test
+using Clustering
+
+@testset "fmeasure()" begin
+
+
+    a1 = [1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 3, 3, 3, 3]
+    a2 = [1, 1, 1, 1, 1, 2, 3, 3, 1, 2, 2, 2, 2, 2, 3, 3, 3]
+    @test fmeasure(a1, a2) ≈ 0.47 atol=1.0e-2
+    @test pair_precision(a1, a2) ≈ 0.5 atol=1.0e-2
+    @test pair_recall(a1, a2) ≈ 0.45 atol=1.0e-2
+
+
+    a1 = [1, 1, 1, 1, 1, 3, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 1, 2]
+    a2 = [1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4]
+    @test fmeasure(a1, a2) ≈ 0.529 atol=1.0e-2
+    @test pair_precision(a1, a2) ≈ 0.6 atol=1.0e-2
+    @test pair_recall(a1, a2) ≈ 0.47 atol=1.0e-2
+
+end
diff --git a/test/runtests.jl b/test/runtests.jl
index 1f9d483a..ab7e35a6 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -19,7 +19,8 @@ tests = ["seeding",
          "hclust",
          "mcl",
          "vmeasure",
-         "mutualinfo"]
+         "mutualinfo",
+         "fmeasure"]
 
 println("Runing tests:")
 for t in tests