From db8d62c47d5df80a64a71d2f6daa4a84b5502c23 Mon Sep 17 00:00:00 2001
From: d-monnet <70266099+d-monnet@users.noreply.github.com>
Date: Tue, 17 Oct 2023 11:35:58 -0400
Subject: [PATCH] add csv export option (#95)

* add csv export option

* Update src/performance_profiles.jl

Co-authored-by: tmigot <tangi.migot@gmail.com>

* Update src/performance_profiles.jl

Co-authored-by: tmigot <tangi.migot@gmail.com>

* Update src/performance_profiles.jl

Co-authored-by: tmigot <tangi.migot@gmail.com>

* Update src/performance_profiles.jl

Co-authored-by: tmigot <tangi.migot@gmail.com>

* update signature
add header length check

* Update documentation

* add test for coverage

* fix doc

---------

Co-authored-by: tmigot <tangi.migot@gmail.com>
---
 Project.toml                |  4 +++
 src/BenchmarkProfiles.jl    |  2 +-
 src/performance_profiles.jl | 56 +++++++++++++++++++++++++++++++++++++
 test/runtests.jl            | 11 ++++++++
 4 files changed, 72 insertions(+), 1 deletion(-)

diff --git a/Project.toml b/Project.toml
index 0cb0465..e5e85da 100644
--- a/Project.toml
+++ b/Project.toml
@@ -3,16 +3,20 @@ uuid = "ecbce9bc-3e5e-569d-9e29-55181f61f8d0"
 version = "0.4.3"
 
 [deps]
+CSV = "336ed68f-0bac-5ca0-87d4-7b16caf5d00b"
 LaTeXStrings = "b964fa9f-0449-5b57-a5c2-d3ea65f4040f"
 NaNMath = "77ba4419-2d1f-58cd-9bb1-8ffee604a2e3"
 Printf = "de0858da-6303-5e67-8744-51eddeeeb8d7"
 Requires = "ae029012-a4dd-5104-9daa-d747884805df"
+Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 
 [compat]
+CSV = "0.10"
 LaTeXStrings = "^1.3"
 NaNMath = "0.3, 1"
 Requires = "1"
 julia = "^1.6"
+Tables = "1.11"
 
 [extras]
 PGFPlotsX = "8314cec4-20b6-5062-9cdb-752b83310925"
diff --git a/src/BenchmarkProfiles.jl b/src/BenchmarkProfiles.jl
index 982f2e6..2668e7d 100644
--- a/src/BenchmarkProfiles.jl
+++ b/src/BenchmarkProfiles.jl
@@ -5,7 +5,7 @@ import NaNMath
 using Requires
 using Printf
 
-export performance_ratios, performance_profile, performance_profile_data
+export performance_ratios, performance_profile, performance_profile_data, export_performance_profile
 export data_ratios, data_profile
 export bp_backends, PlotsBackend, UnicodePlotsBackend, PGFPlotsXBackend
 
diff --git a/src/performance_profiles.jl b/src/performance_profiles.jl
index 9918bd7..a937ab3 100644
--- a/src/performance_profiles.jl
+++ b/src/performance_profiles.jl
@@ -4,6 +4,8 @@
 #
 # D. Orban, 2015, 2016.
 
+using CSV, Tables
+
 """Compute performance ratios used to produce a performance profile.
 
 There is normally no need to call this function directly.
@@ -152,3 +154,57 @@ function performance_profile(
     kwargs...,
   )
 end
+
+"""
+export_performance_profile(T, filename; solver_names = [], header, kwargs...)
+
+Export a performance profile plot data as .csv file. Profiles data are padded with `NaN` to ensure .csv consistency.
+
+## Arguments
+
+* `T :: Matrix{Float64}`: each column of `T` defines the performance data for a solver (smaller is better).
+  Failures on a given problem are represented by a negative value, an infinite value, or `NaN`.
+* `filename :: String` : path to the export file.
+
+## Keyword Arguments
+
+* `solver_names :: Vector{S}` : names of the solvers
+* `header::Vector{String}`: Contains .csv file column names. Note that `header` value does not change columns order in .csv exported files (see Output).
+
+Other keyword arguments are passed `performance_profile_data`.
+
+Output:
+File containing profile data in .csv format. Columns are solver1_x, solver1_y, solver2_x, ...
+"""
+function export_performance_profile(
+  T::Matrix{Float64},
+  filename::String;
+  solver_names::Vector{S} = String[],
+  header::Vector{S} = String[],
+  kwargs...
+) where {S <: AbstractString}
+  nsolvers = size(T)[2]
+
+  x_data, y_data, max_ratio = performance_profile_data(T;kwargs...)
+  max_elem = maximum(length.(x_data))
+  for i in eachindex(x_data)
+    append!(x_data[i],[NaN for i=1:max_elem-length(x_data[i])])
+    append!(y_data[i],[NaN for i=1:max_elem-length(y_data[i])])
+  end
+  x_mat = hcat(x_data...)
+  y_mat = hcat(y_data...)
+
+  isempty(solver_names) && (solver_names = ["solver_$i" for i = 1:nsolvers])
+
+  if !isempty(header)
+    header_l = size(T)[2]*2
+    length(header) == header_l || error("Header should contain $(header_l) elements")
+    header = vcat([[sname*"_x",sname*"_y"] for sname in solver_names]...)
+  end
+  data = Matrix{Float64}(undef,max_elem,nsolvers*2)
+  for i =0:nsolvers-1
+    data[:,2*i+1] .= x_mat[:,i+1]
+    data[:,2*i+2] .= y_mat[:,i+1]
+  end
+  CSV.write(filename,Tables.table(data),header=header)
+end
diff --git a/test/runtests.jl b/test/runtests.jl
index 968302a..67d58b1 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -61,4 +61,15 @@ if !Sys.isfreebsd() # GR_jll not available, so Plots won't install
     profile = data_profile(PGFPlotsXBackend(), H, T, labels)
     @test isa(profile, Plots.Plot)
   end
+
+  @testset "csv export" begin
+    T = 10 * rand(25, 3)
+    filename = "data.csv"
+    export_performance_profile(T,filename)
+    @test isfile(filename)
+    rm(filename)
+    export_performance_profile(T,filename,header = ["" for _=1:size(T,2)*2])
+    @test isfile(filename)
+    rm(filename)
+  end
 end