Makefile and scripts cleanup; lazar, R and tensorflow tables

8 files changed, 19 insertions, 99 deletions

diff --git a/scripts/confusion-matrix2table.rb b/scripts/confusion-matrix2table.rb
deleted file mode 100755
index ccb4817..0000000
--- a/scripts/confusion-matrix2table.rb
+++ /dev/null
@@ -1,10 +0,0 @@
-#!/usr/bin/env ruby
-
-mat = []
-File.readlines(ARGV[0]).each do |l|
-  mat << l.chomp.split(",")
-end
-puts ",,Predictions,"
-puts ",,mutagenic,non-mutagenic"
-puts "Measurements,mutagenic,#{mat[0][0]},#{mat[0][1]}"
-puts ",non-mutagenic,#{mat[1][0]},#{mat[1][1]}"
diff --git a/scripts/crossvalidation-summary.rb b/scripts/crossvalidation-summary.rb
deleted file mode 100755
index 13b0dfa..0000000
--- a/scripts/crossvalidation-summary.rb
+++ /dev/null
@@ -1,9 +0,0 @@
-#!/usr/bin/env ruby
-require_relative '../../lazar/lib/lazar'
-include OpenTox
-
-summary = [] 
-model = Model::Validation.find(File.read(ARGV[0]).chomp).crossvalidations.each do |cv|
-  summary << cv.statistics
-end
-puts JSON.pretty_generate(summary)
diff --git a/scripts/json2csv.rb b/scripts/json2csv.rb
deleted file mode 100755
index 03191de..0000000
--- a/scripts/json2csv.rb
+++ /dev/null
@@ -1,29 +0,0 @@
-#!/usr/bin/env ruby
-require_relative '../../lazar/lib/lazar'
-include OpenTox
-
-results = JSON.parse File.read(ARGV[0])
-puts "SMILES,#{File.basename(ARGV[0],".json").sub("pa_","").capitalize},Probability(0),Probability(1),Nr Neighbors,Warnings"
-results.each do |id,r|
-  s = Compound.find(id).smiles
-	if r["value"]
-		puts [
-			s,
-			r["value"],
-			r["probabilities"]["0"],
-			r["probabilities"]["1"],
-			r["neighbors"].size,
-			r["warnings"],
-		].join(",")
-	else
-		r["neighbors"] ? n = r["neighbors"].size : n = nil
-		puts [
-			s,
-			r["value"],
-			nil,
-			nil,
-			n,
-			r["warnings"],
-		].join(",")
-	end
-end
diff --git a/scripts/results.rb b/scripts/results.rb
deleted file mode 100755
index 1a36278..0000000
--- a/scripts/results.rb
+++ /dev/null
@@ -1,36 +0,0 @@
-#!/usr/bin/env ruby
-require 'json'
-
-result = {}
-ARGV.each do |f|
-  fname = File.basename(f,".json")
-  program,algo = fname.split('-')
-  case program
-  when "tensorflow"
-    algo == "all" ? algo = "without feature selection" : algo = "with feature selection"
-  when "lazar"
-    algo = "high-confidence" if algo == "high"
-    if algo == "padel"
-      algo = "PaDEL" 
-      fname.match("high") ? algo += " high-confidence" : algo += " all"
-    end
-  end
-  result[program] ||= {}
-  result[program][algo] = JSON.parse(File.read(f)).collect{|k,v| [k,v.round(2)]}.to_h
-end
-
-out = {:programs => []}
-result.keys.each do |prog|
-  out[:programs] << {:name => prog, :algos => []}
-  result[prog].keys.each do |algo|
-    r = result[prog][algo].dup
-    result[prog][algo].each do |k,v|
-      r[k+"_perc"] = (v*100).round
-    end
-    r[:name] = algo
-    r[:abbrev] = prog+"-"+algo
-    out[:programs].last[:algos] << r
-  end
-end
-
-puts out.to_json
diff --git a/scripts/roc.R b/scripts/roc.R
index cb219fc..afc8293 100755
--- a/scripts/roc.R
+++ b/scripts/roc.R
@@ -1,6 +1,6 @@
 #!/usr/bin/env Rscript
 library(ggplot2)
-data <- read.csv("figures/results.csv",header=T)
+data <- read.csv("figures/roc.csv",header=T)
 p <- ggplot(data, aes(x=fpr, y=tpr)) + geom_abline()
 p <- p + geom_label(label=rownames(data) )
 p <- p + expand_limits(x=c(0,1),y=c(0,1))
diff --git a/scripts/summaries2table.rb b/scripts/summaries2table.rb
index 5470b26..f98ec54 100755
--- a/scripts/summaries2table.rb
+++ b/scripts/summaries2table.rb
@@ -1,19 +1,23 @@
 #!/usr/bin/env ruby
-require 'json'
+require 'yaml'
 
-results = {}
+rows = {:acc => "Accuracy", :tpr => "True positive rate/Sensitivity", :tnr => "True negative rate/Specificity", :ppv => "Positive predictive value/Precision", :npv => "Negative predictive value", :n => "Nr. predictions"}
+data = YAML.load_file "10-fold-crossvalidations/summary.yaml"
 
-ARGV.each do |f|
-  results[File.basename(f,".json")] = JSON.parse(File.read(f))
+case ARGV[0]
+when "R"
+  header = ["RF","SVM","DL"]
+  keys = header.collect{|h| "R-"+h}
+when "tensorflow"
+  header = ["RF","LR (SGD)","LR (SCIKIT)","NN"]
+  keys = ["lr","lr2","nn"].collect{|n| "tensorflow-"+n+".v3"}
+when "lazar"
+  header = ["lazar-mp2d (all)","lazar-mp2d (high confidence)", "lazar-padel (all)","lazar-padel (high confidence)"]
+  keys = ["lazar-all","lazar-high-confidence", "lazar-padel-all","lazar-padel-high-confidence"]
 end
-
-print ","
-puts results.keys.collect{|k| k.sub("tensorflow","TF")}.join(",")
-["accuracy","true_positive_rate","true_negative_rate","positive_predictive_value","negative_predictive_value"].each do |m|
-  line = [m.gsub("_"," ")]
-  results.each do |k,v|
-    line << v[m].round(2)
-  end
-  puts line.join(",")
+puts ","+header.join(",")
+rows.each do |short,long|
+  print long+","
+  puts keys.collect{|k| data[k][short]}.join(",")
 end
-
+exit
diff --git a/scripts/mp2d-tsne.R b/scripts/tsne-mp2d.R
index 0877622..0877622 100755
--- a/scripts/mp2d-tsne.R
+++ b/scripts/tsne-mp2d.R
diff --git a/scripts/padel-tsne.R b/scripts/tsne-padel.R
index b8e9763..b8e9763 100755
--- a/scripts/padel-tsne.R
+++ b/scripts/tsne-padel.R