Backup including important chances from devjl_dv

train_succeess and dataset cleanup + own chances (get compound name
function, small dataset fix)

4 files changed, 52 insertions, 8 deletions

diff --git a/lib/algorithm.rb b/lib/algorithm.rb
index 40cbc84..19666ad 100644
--- a/lib/algorithm.rb
+++ b/lib/algorithm.rb
@@ -381,10 +381,10 @@ module OpenTox
         else
           #LOGGER.debug gram_matrix.to_yaml
           @r = RinRuby.new(false,false) # global R instance leads to Socket errors after a large number of requests
-          @r.eval "set.seed(1)"
           @r.eval "suppressPackageStartupMessages(library('caret'))" # requires R packages "caret" and "kernlab"
           @r.eval "suppressPackageStartupMessages(library('doMC'))" # requires R packages "multicore"
           @r.eval "registerDoMC()" # switch on parallel processing
+          @r.eval "set.seed(1)"
           begin
 
             # set data
@@ -417,7 +417,7 @@ module OpenTox
 
             # model + support vectors
             LOGGER.debug "Creating R SVM model ..."
-            @r.eval <<-EOR
+            train_success = @r.eval <<-EOR
               model = train(prop_matrix,y,method="svmradial",tuneLength=8,trControl=trainControl(method="LGOCV",number=10),preProcess=c("center", "scale"))
               perf = ifelse ( class(y)!='numeric', max(model$results$Accuracy), model$results[which.min(model$results$RMSE),]$Rsquared )
             EOR
@@ -431,6 +431,7 @@ module OpenTox
 
             # censoring
             prediction = nil if ( @r.perf.nan? || @r.perf < min_train_performance )
+            prediction = nil unless train_success
             LOGGER.debug "Performance: #{sprintf("%.2f", @r.perf)}"
           rescue Exception => e
             LOGGER.debug "#{e.class}: #{e.message}"
@@ -459,12 +460,12 @@ module OpenTox
         
         # need packs 'randomForest', 'RANN'
         @r.eval <<-EOR
-          set.seed(1)
           suppressPackageStartupMessages(library('caret'))
           suppressPackageStartupMessages(library('randomForest'))
           suppressPackageStartupMessages(library('RANN'))
           suppressPackageStartupMessages(library('doMC'))
           registerDoMC()
+          set.seed(1)
           
           acts = read.csv(ds_csv_file, check.names=F)
           feats = read.csv(fds_csv_file, check.names=F)
@@ -506,15 +507,16 @@ module OpenTox
           features = features[,!names(features) %in% nan_col]
          
           # determine subsets
-          subsets = dim(features)[2]*c(0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7)
-          subsets = c(2,3,4,5,7,10,subsets)
+          subsets = dim(features)[2]*c(0.3, 0.32, 0.34, 0.36, 0.38, 0.4, 0.42, 0.44, 0.46, 0.48, 0.5, 0.52, 0.54, 0.56, 0.58, 0.6, 0.62, 0.64, 0.66, 0.68, 0.7)
+          #subsets = dim(features)[2]*c(0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7)
+          #subsets = c(2,3,4,5,7,10,subsets)
           #subsets = c(2,3,4,5,7,10,13,16,19,22,25,28,30)
           subsets = unique(sort(round(subsets))) 
           subsets = subsets[subsets<=dim(features)[2]]
           subsets = subsets[subsets>1] 
          
           # Recursive feature elimination
-          rfProfile = rfe( x=features, y=y, rfeControl=rfeControl(functions=rfFuncs, method='cv'), sizes=subsets)
+          rfProfile = rfe( x=features, y=y, rfeControl=rfeControl(functions=rfFuncs, number=150), sizes=subsets)
           optVar = rfProfile$optVariables
           if (rfProfile$bestSubset == dim(features)[2]) {
             newRMSE = rfProfile$results$RMSE
diff --git a/lib/compound.rb b/lib/compound.rb
index c7c82c3..5024af1 100644
--- a/lib/compound.rb
+++ b/lib/compound.rb
@@ -3,6 +3,7 @@
 
 module OpenTox
 
+  require "rexml/document"
   # Ruby wrapper for OpenTox Compound Webservices (http://opentox.org/dev/apis/api-1.2/structure).
 	class Compound 
 
@@ -130,6 +131,47 @@ module OpenTox
         "not available"
       end
 		end
+    
+    
+    # Get all known compound names sorted by classification. Relies on an external service for name lookups.
+    # @example
+    #   names = compound.to_names_hash
+    # @return [Hash] Classification => Name Array
+		def to_names_hash
+      begin
+        xml = RestClientWrapper.get("#{@@cactus_uri}#{@inchi}/names/xml")
+        xmldoc = REXML::Document.new(xml)
+        data = {}
+        
+        xmldoc.root.elements[1].elements.each{|e|
+          if data.has_key?(e.attribute("classification").value) == false
+             data[e.attribute("classification").value] = [e.text]
+          else
+             data[e.attribute("classification").value].push(e.text)
+          end
+        }
+        data
+      rescue
+        "not available"
+      end
+		end
+
+    # Get all known compound names sorted by classification. Relies on an external service for name lookups.
+    # @example
+    #   names = compound.to_names_hash
+    # @return [Hash] Classification => Name Array
+    def to_ambit_names_hash
+      begin
+        ds = OpenTox::Dataset.new
+        ds.save
+        ds.load_rdfxml(RestClientWrapper.get("http://apps.ideaconsult.net:8080/ambit2/query/compound/search/names?type=smiles&property=&search=#{@inchi}"))
+        ds.save
+        ds.uri
+      rescue
+        "not available"
+      end
+    end
+
 
 		# Match a smarts string
     # @example
diff --git a/lib/transform.rb b/lib/transform.rb
index 8fe1093..cb530a3 100644
--- a/lib/transform.rb
+++ b/lib/transform.rb
@@ -396,7 +396,7 @@ module OpenTox
             @q_prop = gsl_q_prop_orig.row(0).to_a
           end
 
-          LOGGER.debug "F: #{@n_prop.size}x#{@n_prop[0].size}; R: #{@q_prop.size}"
+          LOGGER.debug "F: #{@n_prop.size}x#{@n_prop[0].size}; R: #{@q_prop.size}" if (@n_prop && @n_prop[0] && @q_prop)
           LOGGER.debug "Sims: #{@sims.size}, Acts: #{@acts.size}"
 
           @sims = [ gram_matrix, @sims ] 
diff --git a/lib/utils.rb b/lib/utils.rb
index d96c2b3..eccec46 100644
--- a/lib/utils.rb
+++ b/lib/utils.rb
@@ -60,7 +60,7 @@ module OpenTox
         ds = OpenTox::Dataset.new
         ds.save
         parser.dataset = ds
-        if compounds.size == 1
+        if compounds.size < 4 
           ds = parser.load_csv(master.collect{|r| r.join(",")}.join("\n"),false,true,false) # all_numeric = true, all features should be treated as numeric
         else
           ds = parser.load_csv(master.collect{|r| r.join(",")}.join("\n"),false,false,true) # del_nominal = true, which removes nominal features