add selected results

Author: flixpar

selected_results/csv_to_table.jl

@@ -0,0 +1,48 @@
+using CSV
+using DataFrames
+
+fn = ARGS[1]
+
+data = DataFrame(CSV.File(fn))
+
+constcols = [c for c in names(data) if all(coalesce.(data[!,c],"") .== coalesce(data[1,c],""))]
+constvals = [c => data[1,c] for c in constcols]
+select!(data, Not(constcols))
+
+table = String[]
+
+colname_cvt = Dict("N" => "N", "method" => "Method", "time" => "Time", "obj" => "Objective", "ratio" => "\$\\alpha\$")
+header = [haskey(colname_cvt, c) ? colname_cvt[c] : c for c in names(data)]
+header_str = foldl((a,b) -> a * " & " * b, header) * " \\\\\n"
+push!(table, header_str)
+push!(table, "\\hline\n")
+
+method_lookup = Dict(
+	"opt" => "Quadratic Programming",
+	"lin" => "Linearization",
+	"lpr" => "LP + Rounding",
+	"faq" => "FAQ",
+	"rand" => "Random Search",
+	"mms" => "MMS",
+	"ns" => "NS",
+	"star" => "Common Star Packing",
+	"dense" => "Dense Subgraph Mapping",
+)
+
+fmt(v::String, m) = (m == "method") ? method_lookup[v] : v
+fmt(v::Int, m) = string(v)
+fmt(v::Float64, m) = (v < 100) ? string(round(v, sigdigits=4)) : string(round(v, digits=2))
+
+for row in eachrow(data)
+	s = [fmt(row[c],c) for c in names(data)]
+	line = foldl((a,b) -> a * " & " * b, s) * " \\\\\n"
+	push!(table, line)
+end
+
+table_str = foldl(*, table)
+
+println("Params:")
+println(constvals)
+
+println("Table:")
+print(table_str)

selected_results/faq_scale.csv

@@ -0,0 +1,11 @@
+objective,matrix_type,density,N,method,time,obj,ratio
+max,known,,10,faq,6.71E-05,96421.25,1
+max,known,,20,faq,0.000256138,312851,1
+max,known,,30,faq,0.000787254,1.48E+06,1
+max,known,,40,faq,0.001605326,1.15E+06,1
+max,known,,50,faq,0.003028191,1.84E+06,1
+max,known,,60,faq,0.004772163,3.92E+06,1
+max,known,,70,faq,0.005355144,3.71E+06,1
+max,known,,80,faq,0.008521134,5.77E+06,1
+max,known,,90,faq,0.012298533,6.68E+06,1
+max,known,,100,faq,0.019067918,1.03E+07,1

selected_results/figs.ipynb

@@ -0,0 +1,6 @@
+objective,matrix_type,density,N,method,time,obj,ratio
+max,known,,5,mms,0.0741457475,9945.5,0.9419275354202834
+max,known,,10,mms,0.8156179415,78180.5,0.7128157250318241
+max,known,,15,mms,9.147652628,75173.0,0.7499126258610194
+max,known,,20,mms,76.90932082,153752.5,0.7334346789476077
+max,known,,25,mms,600.9482155065,396552.0,0.7560127913267869

selected_results/mms_scale.csv

@@ -0,0 +1,8 @@
+N,objective,matrix_type,density,method,time,obj,ratio
+10,max,known,,opt,129.05,93419.4,1
+10,max,known,,lin,22.38,93419.4,1
+10,max,known,,lpr,0.905536152,93419.4,1
+10,max,known,,faq,9.98E-05,93397.8,0.989570256
+10,max,known,,rand,7.03E-05,77525,0.812765734
+10,max,known,,mms,0.887694738,71807.7,0.746288094
+10,max,known,,ns,10.87548241,77917.1,0.813119239

selected_results/n10.csv

@@ -0,0 +1,6 @@
+N,objective,matrix_type,density,method,time,obj,ratio
+15,max,known,,lpr,9.015414592700001,215869.2,1.0
+15,max,known,,faq,0.00016191499999999997,215869.2,1.0
+15,max,known,,rand,0.0028081698,172240.9,0.7944852890316068
+15,max,known,,mms,9.3054625113,166797.6,0.7641117394933113
+15,max,known,,ns,201.57749361880002,170181.0,0.7870023541726983

selected_results/n15.csv

@@ -0,0 +1,6 @@
+N,objective,matrix_type,density,method,time,obj,ratio
+20,max,known,,lpr,75.96571031330001,383417.8,1.0
+20,max,known,,faq,0.0002616782,383417.8,1.0
+20,max,known,,rand,0.00041320230000000004,297737.5,0.7765360540786597
+20,max,known,,mms,76.624266053,288392.4,0.7540964317351115
+20,max,known,,ns,1937.445926107,296620.4,0.7694145853951608

selected_results/n20.csv

@@ -0,0 +1,6 @@
+objective,matrix_type,density,N,method,time,obj,ratio
+max,known,,5,ns,0.10717362450000001,8572.0,0.9258713136729222
+max,known,,10,ns,6.471749943000001,71237.5,0.7802216676016591
+max,known,,15,ns,79.7950929615,249840.0,0.7800719082550769
+max,known,,20,ns,500.06151821049997,480764.5,0.7731621451136911
+max,known,,25,ns,1932.1197390430002,284943.0,0.7744593799192648