From 3e7d98147ef55e4e6e55a518a05c6eecee6f769b Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Thu, 19 Jan 2017 12:41:55 +0000
Subject: fixed feature test and unique_descriptors

---
 lib/unique_descriptors.rb |  2 +-
 test/feature.rb           | 10 +++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/lib/unique_descriptors.rb b/lib/unique_descriptors.rb
index 8341a67..1b19d6a 100644
--- a/lib/unique_descriptors.rb
+++ b/lib/unique_descriptors.rb
@@ -48,7 +48,7 @@ UNIQUEDESCRIPTORS = [
   #"Cdk.HBondAcceptorCount", #Descriptor that calculates the number of hydrogen bond acceptors.
   #"Cdk.HBondDonorCount", #Descriptor that calculates the number of hydrogen bond donors.
   "Cdk.HybridizationRatio", #Characterizes molecular complexity in terms of carbon hybridization states.
-  "Cdk.IPMolecularLearning", #Descriptor that evaluates the ionization potential.
+  #"Cdk.IPMolecularLearning", #Descriptor that evaluates the ionization potential.
   "Cdk.KappaShapeIndices", #Descriptor that calculates Kier and Hall kappa molecular shape indices.
   "Cdk.KierHallSmarts", #Counts the number of occurrences of the E-state fragments
   "Cdk.LargestChain", #Returns the number of atoms in the largest chain
diff --git a/test/feature.rb b/test/feature.rb
index 40edb9f..85ce588 100644
--- a/test/feature.rb
+++ b/test/feature.rb
@@ -57,20 +57,20 @@ class FeatureTest < MiniTest::Test
   def test_physchem_description
     assert_equal 346, PhysChem.descriptors.size
     assert_equal 15, PhysChem.openbabel_descriptors.size
-    assert_equal 295, PhysChem.cdk_descriptors.size
+    assert_equal 286, PhysChem.cdk_descriptors.size
     assert_equal 45, PhysChem.joelib_descriptors.size
-    assert_equal 310, PhysChem.unique_descriptors.size
+    assert_equal 309, PhysChem.unique_descriptors.size
   end
 
   def test_physchem
     assert_equal 346, PhysChem.descriptors.size
     c = Compound.from_smiles "CC(=O)CC(C)C"
     logP = PhysChem.find_or_create_by :name => "Openbabel.logP"
-    assert_equal 1.6215, logP.calculate(c)
+    assert_equal 1.6215, c.calculate_properties([logP]).first
     jlogP = PhysChem.find_or_create_by :name => "Joelib.LogP"
-    assert_equal 3.5951, jlogP.calculate(c)
+    assert_equal 3.5951, c.calculate_properties([jlogP]).first
     alogP = PhysChem.find_or_create_by :name => "Cdk.ALOGP.ALogP"
-    assert_equal 0.35380000000000034, alogP.calculate(c)
+    assert_equal 0.35380000000000034, c.calculate_properties([alogP]).first
   end
 
 end
-- 
cgit v1.2.3


From 4c791043366275f6748a706dd5f73260d0560d41 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Thu, 19 Jan 2017 16:39:46 +0000
Subject: added TODO for missing descriptor

---
 lib/unique_descriptors.rb | 1 +
 1 file changed, 1 insertion(+)

diff --git a/lib/unique_descriptors.rb b/lib/unique_descriptors.rb
index 1b19d6a..fc10cd4 100644
--- a/lib/unique_descriptors.rb
+++ b/lib/unique_descriptors.rb
@@ -48,6 +48,7 @@ UNIQUEDESCRIPTORS = [
   #"Cdk.HBondAcceptorCount", #Descriptor that calculates the number of hydrogen bond acceptors.
   #"Cdk.HBondDonorCount", #Descriptor that calculates the number of hydrogen bond donors.
   "Cdk.HybridizationRatio", #Characterizes molecular complexity in terms of carbon hybridization states.
+  # TODO check why the next descriptor is not present in the CDK_DESCRIPTIONS variable.
   #"Cdk.IPMolecularLearning", #Descriptor that evaluates the ionization potential.
   "Cdk.KappaShapeIndices", #Descriptor that calculates Kier and Hall kappa molecular shape indices.
   "Cdk.KierHallSmarts", #Counts the number of occurrences of the E-state fragments
-- 
cgit v1.2.3


From 4e0088110a6a92fc2787c33c1b09ea76dae70dae Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Thu, 16 Mar 2017 12:53:23 +0000
Subject: loael edit

---
 lib/caret.rb      | 11 +++++------
 lib/dataset.rb    |  2 +-
 lib/model.rb      | 35 ++++++++++++++++++++++-------------
 lib/regression.rb |  2 +-
 4 files changed, 29 insertions(+), 21 deletions(-)

diff --git a/lib/caret.rb b/lib/caret.rb
index f5c2bde..8bccf74 100644
--- a/lib/caret.rb
+++ b/lib/caret.rb
@@ -22,12 +22,11 @@ module OpenTox
         end
         if independent_variables.flatten.uniq == ["NA"] or independent_variables.flatten.uniq == [] 
           prediction = Algorithm::Regression::weighted_average dependent_variables:dependent_variables, weights:weights
-          prediction[:warning] = "No variables for regression model. Using weighted average of similar substances."
+          prediction[:warnings] << "No variables for regression model. Using weighted average of similar substances."
         elsif
           dependent_variables.size < 3
           prediction = Algorithm::Regression::weighted_average dependent_variables:dependent_variables, weights:weights
-          prediction[:warning] = "Insufficient number of neighbors (#{dependent_variables.size}) for regression model. Using weighted average of similar substances."
-
+          prediction[:warnings] << "Insufficient number of neighbors (#{dependent_variables.size}) for regression model. Using weighted average of similar substances."
         else
           dependent_variables.each_with_index do |v,i| 
             dependent_variables[i] = to_r(v)
@@ -52,7 +51,7 @@ module OpenTox
             $logger.debug dependent_variables
             $logger.debug independent_variables
             prediction = Algorithm::Regression::weighted_average dependent_variables:dependent_variables, weights:weights
-            prediction[:warning] = "R caret model creation error. Using weighted average of similar substances."
+            prediction[:warnings] << "R caret model creation error. Using weighted average of similar substances."
             return prediction
           end
           begin
@@ -73,12 +72,12 @@ module OpenTox
             $logger.debug "R caret prediction error for:"
             $logger.debug self.inspect
             prediction = Algorithm::Regression::weighted_average dependent_variables:dependent_variables, weights:weights
-            prediction[:warning] = "R caret prediction error. Using weighted average of similar substances"
+            prediction[:warnings] << "R caret prediction error. Using weighted average of similar substances"
             return prediction
           end
           if prediction.nil? or prediction[:value].nil?
             prediction = Algorithm::Regression::weighted_average dependent_variables:dependent_variables, weights:weights
-            prediction[:warning] = "Could not create local caret model. Using weighted average of similar substances."
+            prediction[:warnings] << "Empty R caret prediction. Using weighted average of similar substances."
           end
         end
         prediction
diff --git a/lib/dataset.rb b/lib/dataset.rb
index 44690e1..6e7d67f 100644
--- a/lib/dataset.rb
+++ b/lib/dataset.rb
@@ -46,7 +46,7 @@ module OpenTox
       if data_entries[substance.to_s] and data_entries[substance.to_s][feature.to_s]
         data_entries[substance.to_s][feature.to_s]
       else
-        nil
+        [nil]
       end
     end
 
diff --git a/lib/model.rb b/lib/model.rb
index b18610d..56d8665 100644
--- a/lib/model.rb
+++ b/lib/model.rb
@@ -57,7 +57,7 @@ module OpenTox
           model.version = {:warning => "git is not installed"}
         end
 
-        # set defaults
+        # set defaults#
         substance_classes = training_dataset.substances.collect{|s| s.class.to_s}.uniq
         bad_request_error "Cannot create models for mixed substance classes '#{substance_classes.join ', '}'." unless substance_classes.size == 1
 
@@ -70,7 +70,7 @@ module OpenTox
             },
             :similarity => {
               :method => "Algorithm::Similarity.tanimoto",
-              :min => 0.1
+              :min => 0.5,
             },
             :feature_selection => nil
           }
@@ -81,7 +81,7 @@ module OpenTox
             }
           elsif model.class == LazarRegression
             model.algorithms[:prediction] = {
-              :method => "Algorithm::Caret.pls",
+              :method => "Algorithm::Caret.rf",
             }
           end
 
@@ -93,7 +93,7 @@ module OpenTox
             },
             :similarity => {
               :method => "Algorithm::Similarity.weighted_cosine",
-              :min => 0.5
+              :min => 0.5,
             },
             :prediction => {
               :method => "Algorithm::Caret.rf",
@@ -191,7 +191,7 @@ module OpenTox
       # Predict a substance (compound or nanoparticle)
       # @param [OpenTox::Substance]
       # @return [Hash]
-      def predict_substance substance
+      def predict_substance substance, threshold = self.algorithms[:similarity][:min]
         
         @independent_variables = Marshal.load $gridfs.find_one(_id: self.independent_variables_id).data
         case algorithms[:similarity][:method]
@@ -221,20 +221,19 @@ module OpenTox
           bad_request_error "Unknown descriptor type '#{descriptors}' for similarity method '#{similarity[:method]}'."
         end
         
-        prediction = {}
+        prediction = {:warnings => [], :measurements => []}
+        prediction[:warnings] << "Similarity threshold #{threshold} < #{algorithms[:similarity][:min]}, prediction may be out of applicability domain." if threshold < algorithms[:similarity][:min]
         neighbor_ids = []
         neighbor_similarities = []
         neighbor_dependent_variables = []
         neighbor_independent_variables = []
 
-        prediction = {}
         # find neighbors
         substance_ids.each_with_index do |s,i|
           # handle query substance
           if substance.id.to_s == s
-            prediction[:measurements] ||= []
             prediction[:measurements] << dependent_variables[i]
-            prediction[:warning] = "Substance '#{substance.name}, id:#{substance.id}' has been excluded from neighbors, because it is identical with the query substance."
+            prediction[:info] = "Substance '#{substance.name}, id:#{substance.id}' has been excluded from neighbors, because it is identical with the query substance."
           else
             if fingerprints?
               neighbor_descriptors = fingerprints[i]
@@ -243,7 +242,7 @@ module OpenTox
               neighbor_descriptors = scaled_variables.collect{|v| v[i]}
             end
             sim = Algorithm.run algorithms[:similarity][:method], [similarity_descriptors, neighbor_descriptors, descriptor_weights]
-            if sim >= algorithms[:similarity][:min]
+            if sim >= threshold
               neighbor_ids << s
               neighbor_similarities << sim
               neighbor_dependent_variables << dependent_variables[i]
@@ -258,17 +257,27 @@ module OpenTox
         measurements = nil
         
         if neighbor_similarities.empty?
-          prediction.merge!({:value => nil,:warning => "Could not find similar substances with experimental data in the training dataset.",:neighbors => []})
+          prediction[:value] = nil
+          prediction[:warnings] << "Could not find similar substances with experimental data in the training dataset."
         elsif neighbor_similarities.size == 1
-          prediction.merge!({:value => dependent_variables.first, :probabilities => nil, :warning => "Only one similar compound in the training set. Predicting its experimental value.", :neighbors => [{:id => neighbor_ids.first, :similarity => neighbor_similarities.first}]})
+          prediction[:value] = nil
+          prediction[:warnings] << "Cannot create prediction: Only one similar compound in the training set."
+          prediction[:neighbors] = [{:id => neighbor_ids.first, :similarity => neighbor_similarities.first}]
         else
           query_descriptors.collect!{|d| d ? 1 : 0} if algorithms[:feature_selection] and algorithms[:descriptors][:method] == "fingerprint"
           # call prediction algorithm
           result = Algorithm.run algorithms[:prediction][:method], dependent_variables:neighbor_dependent_variables,independent_variables:neighbor_independent_variables ,weights:neighbor_similarities, query_variables:query_descriptors
           prediction.merge! result
           prediction[:neighbors] = neighbor_ids.collect_with_index{|id,i| {:id => id, :measurement => neighbor_dependent_variables[i], :similarity => neighbor_similarities[i]}}
+          #if neighbor_similarities.max < algorithms[:similarity][:warn_min]
+            #prediction[:warnings] << "Closest neighbor has similarity < #{algorithms[:similarity][:warn_min]}. Prediction may be out of applicability domain."
+          #end
+        end
+        if prediction[:warnings].empty? or threshold < algorithms[:similarity][:min]
+          prediction
+        else # try again with a lower threshold
+          predict_substance substance, 0.2
         end
-        prediction
       end
 
       # Predict a substance (compound or nanoparticle), an array of substances or a dataset
diff --git a/lib/regression.rb b/lib/regression.rb
index fd2855f..25c0732 100644
--- a/lib/regression.rb
+++ b/lib/regression.rb
@@ -17,7 +17,7 @@ module OpenTox
           sim_sum += weights[i]
         end if dependent_variables
         sim_sum == 0 ? prediction = nil : prediction = weighted_sum/sim_sum
-        {:value => prediction}
+        {:value => prediction, :warnings => ["Weighted average prediction, no prediction interval available."]}
       end
 
     end
-- 
cgit v1.2.3


From a8e6ea94d037f576be699b43751702ef5bc37496 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Tue, 21 Mar 2017 09:18:49 +0000
Subject: fixed test_to_csv

---
 test/dataset.rb | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/dataset.rb b/test/dataset.rb
index e91e65a..055a029 100644
--- a/test/dataset.rb
+++ b/test/dataset.rb
@@ -160,7 +160,7 @@ class DatasetTest < MiniTest::Test
         if v.numeric?
           assert_equal v.to_f, serialized[inchi][i].to_f
         else
-          assert_equal v, serialized[inchi][i]
+          assert_equal v.to_s, serialized[inchi][i].to_s
         end
       end
 
-- 
cgit v1.2.3


From 5d7aec4c09709f2179bbbac1e1140255156c0cda Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Tue, 21 Mar 2017 15:08:38 +0000
Subject: fixed endless loop; rescue for missing neighbors

---
 lib/model.rb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/model.rb b/lib/model.rb
index 56d8665..7cc6765 100644
--- a/lib/model.rb
+++ b/lib/model.rb
@@ -273,7 +273,7 @@ module OpenTox
             #prediction[:warnings] << "Closest neighbor has similarity < #{algorithms[:similarity][:warn_min]}. Prediction may be out of applicability domain."
           #end
         end
-        if prediction[:warnings].empty? or threshold < algorithms[:similarity][:min]
+        if prediction[:warnings].empty? or threshold <= algorithms[:similarity][:min]
           prediction
         else # try again with a lower threshold
           predict_substance substance, 0.2
@@ -309,7 +309,7 @@ module OpenTox
         # serialize result
         if object.is_a? Substance
           prediction = predictions[substances.first.id.to_s]
-          prediction[:neighbors].sort!{|a,b| b[1] <=> a[1]} # sort according to similarity
+          prediction[:neighbors].sort!{|a,b| b[1] <=> a[1]} if prediction[:neighbors]# sort according to similarity
           return prediction
         elsif object.is_a? Array
           return predictions
-- 
cgit v1.2.3


From 658cee5e1df2e1fcb0c7f9259955f1e3199deb5a Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Tue, 21 Mar 2017 15:11:44 +0000
Subject: fixed regression test for rf algorithm see: 'loael edit' commit;along
 with larger treining datasets

---
 test/model-regression.rb | 28 ++++++++++++++--------------
 test/setup.rb            |  2 ++
 2 files changed, 16 insertions(+), 14 deletions(-)

diff --git a/test/model-regression.rb b/test/model-regression.rb
index 86b927c..5903e88 100644
--- a/test/model-regression.rb
+++ b/test/model-regression.rb
@@ -10,21 +10,21 @@ class LazarRegressionTest < MiniTest::Test
       },
       :similarity => {
         :method => "Algorithm::Similarity.tanimoto",
-        :min => 0.1
+        :min => 0.5
       },
       :prediction => {
-        :method => "Algorithm::Caret.pls",
+        :method => "Algorithm::Caret.rf",
       },
       :feature_selection => nil,
     }
-    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.medi_log10.csv")
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM_log10.csv")
     model = Model::Lazar.create  training_dataset: training_dataset
     assert_kind_of Model::LazarRegression, model
     assert_equal algorithms, model.algorithms
-    substance = training_dataset.substances[10]
+    substance = training_dataset.substances[145]
     prediction = model.predict substance
     assert_includes prediction[:prediction_interval][0]..prediction[:prediction_interval][1], prediction[:measurements].median, "This assertion assures that measured values are within the prediction interval. It may fail in 5% of the predictions."
-    substance = Compound.from_smiles "NC(=O)OCCC"
+    substance = Compound.from_smiles "c1ccc(cc1)Oc1ccccc1"
     prediction = model.predict substance
     refute_nil prediction[:value]
     refute_nil prediction[:prediction_interval]
@@ -59,8 +59,8 @@ class LazarRegressionTest < MiniTest::Test
     model = Model::Lazar.create training_dataset: training_dataset, algorithms: algorithms
     compound = Compound.from_smiles "CCCSCCSCC"
     prediction = model.predict compound
-    assert_equal 4, prediction[:neighbors].size
-    assert_equal 1.37, prediction[:value].round(2)
+    assert_equal 3, prediction[:neighbors].size
+    assert prediction[:value].round(2) > 1.37, "Prediction value (#{prediction[:value].round(2)}) should be larger than 1.37."
   end
 
   def test_local_physchem_regression
@@ -112,12 +112,12 @@ class LazarRegressionTest < MiniTest::Test
         :method => "Algorithm::Similarity.cosine",
       }
     }
-    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.mini_log10.csv")
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.medi_log10.csv")
     model = Model::Lazar.create  training_dataset: training_dataset, algorithms: algorithms
     assert_kind_of Model::LazarRegression, model
-    assert_equal "Algorithm::Caret.pls", model.algorithms[:prediction][:method]
+    assert_equal "Algorithm::Caret.rf", model.algorithms[:prediction][:method]
     assert_equal "Algorithm::Similarity.cosine", model.algorithms[:similarity][:method]
-    assert_equal 0.1, model.algorithms[:similarity][:min]
+    assert_equal 0.5, model.algorithms[:similarity][:min]
     algorithms[:descriptors].delete :features
     assert_equal algorithms[:descriptors], model.algorithms[:descriptors]
     prediction = model.predict training_dataset.substances[10]
@@ -130,14 +130,14 @@ class LazarRegressionTest < MiniTest::Test
         :method => "Algorithm::FeatureSelection.correlation_filter",
       },
     }
-    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.mini_log10.csv")
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM_log10.csv")
     model = Model::Lazar.create  training_dataset: training_dataset, algorithms: algorithms
     assert_kind_of Model::LazarRegression, model
-    assert_equal "Algorithm::Caret.pls", model.algorithms[:prediction][:method]
+    assert_equal "Algorithm::Caret.rf", model.algorithms[:prediction][:method]
     assert_equal "Algorithm::Similarity.tanimoto", model.algorithms[:similarity][:method]
-    assert_equal 0.1, model.algorithms[:similarity][:min]
+    assert_equal 0.5, model.algorithms[:similarity][:min]
     assert_equal algorithms[:feature_selection][:method], model.algorithms[:feature_selection][:method]
-    prediction = model.predict training_dataset.substances[10]
+    prediction = model.predict training_dataset.substances[145]
     refute_nil prediction[:value]
   end
 
diff --git a/test/setup.rb b/test/setup.rb
index 40c8ebf..c1cddfb 100644
--- a/test/setup.rb
+++ b/test/setup.rb
@@ -3,6 +3,8 @@ require 'minitest/autorun'
 require_relative '../lib/lazar.rb'
 #require 'lazar'
 include OpenTox
+#$mongo.database.drop
+#$gridfs = $mongo.database.fs # recreate GridFS indexes
 TEST_DIR ||= File.expand_path(File.dirname(__FILE__))
 DATA_DIR ||= File.join(TEST_DIR,"data")
 training_dataset = Dataset.where(:name => "Protein Corona Fingerprinting Predicts the Cellular Interaction of Gold and Silver Nanoparticles").first
-- 
cgit v1.2.3


From 2cebe2c0dc60bda0c94d0d9c38e7931efadd7155 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Fri, 24 Mar 2017 12:51:12 +0000
Subject: adjusted for higher similarity min value

---
 test/model-classification.rb | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/test/model-classification.rb b/test/model-classification.rb
index 1424f6a..0bb3e83 100644
--- a/test/model-classification.rb
+++ b/test/model-classification.rb
@@ -10,35 +10,35 @@ class LazarClassificationTest < MiniTest::Test
       },
       :similarity => {
         :method => "Algorithm::Similarity.tanimoto",
-        :min => 0.1
+        :min => 0.5
       },
+      :feature_selection => nil,
       :prediction => {
         :method => "Algorithm::Classification.weighted_majority_vote",
       },
-      :feature_selection => nil,
     }
     training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"hamster_carcinogenicity.csv")
     model = Model::Lazar.create  training_dataset: training_dataset
     assert_kind_of Model::LazarClassification, model
     assert_equal algorithms, model.algorithms
-    substance = training_dataset.substances[10]
+    substance = training_dataset.substances[49]
     prediction = model.predict substance
     assert_equal "false", prediction[:value]
     [ {
-      :compound => OpenTox::Compound.from_inchi("InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"),
+      :compound => OpenTox::Compound.from_inchi("InChI=1S/C6H14N2O4/c1-5(10)2-8(7-12)3-6(11)4-9/h5-6,9-11H,2-4H2,1H3"),
       :prediction => "false",
     },{
-      :compound => OpenTox::Compound.from_smiles("c1ccccc1NN"),
+      :compound => OpenTox::Compound.from_smiles("OCC(CN(CC(O)C)N=O)O"),
       :prediction => "false",
     } ].each do |example|
       prediction = model.predict example[:compound]
       assert_equal example[:prediction], prediction[:value]
     end
 
-    compound = Compound.from_smiles "CCO"
+    compound = Compound.from_smiles "O=NN1CCC1"
     prediction = model.predict compound
     assert_equal "true", prediction[:value]
-    assert_equal ["false"], prediction[:measurements]
+    #assert_equal ["false"], prediction[:measurements]
 
     # make a dataset prediction
     compound_dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.mini_log10.csv")
@@ -46,12 +46,12 @@ class LazarClassificationTest < MiniTest::Test
     assert_equal compound_dataset.compounds, prediction_dataset.compounds
 
     cid = prediction_dataset.compounds[7].id.to_s
-    assert_equal "Could not find similar substances with experimental data in the training dataset.", prediction_dataset.predictions[cid][:warning]
+    assert_equal "Could not find similar substances with experimental data in the training dataset.", prediction_dataset.predictions[cid][:warnings][0]
     prediction_dataset.predictions.each do |cid,pred|
-      assert_equal "Could not find similar substances with experimental data in the training dataset.", pred[:warning] if pred[:value].nil?
+      assert_equal "Could not find similar substances with experimental data in the training dataset.", pred[:warnings][0] if pred[:value].nil?
     end
     cid = Compound.from_smiles("CCOC(=O)N").id.to_s
-    assert_match "excluded", prediction_dataset.predictions[cid][:warning]
+    assert_match "excluded", prediction_dataset.predictions[cid][:info]
     # cleanup
     [training_dataset,model,compound_dataset,prediction_dataset].each{|o| o.delete}
   end
@@ -85,7 +85,7 @@ class LazarClassificationTest < MiniTest::Test
     model = Model::Lazar.create training_dataset: training_dataset
     t = Time.now
     2.times do
-      compound = Compound.from_smiles("Clc1ccccc1NN")
+      compound = Compound.from_smiles("OCC(CN(CC(O)C)N=O)O")
       prediction = model.predict compound
       assert_equal "1", prediction[:value]
     end
-- 
cgit v1.2.3


From 04f1f35f2248db65b313c382ee57a53047778472 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Fri, 24 Mar 2017 13:00:19 +0000
Subject: adjusted for higher default similarity min value

---
 test/validation-classification.rb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/test/validation-classification.rb b/test/validation-classification.rb
index fb4c3e7..ac25b29 100644
--- a/test/validation-classification.rb
+++ b/test/validation-classification.rb
@@ -47,9 +47,9 @@ class ValidationClassificationTest < MiniTest::Test
     dataset = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
     model = Model::Lazar.create training_dataset: dataset
     loo = ClassificationLeaveOneOut.create model
-    assert_equal 14, loo.nr_unpredicted
+    assert_equal 77, loo.nr_unpredicted
     refute_empty loo.confusion_matrix
-    assert loo.accuracy > 0.77
+    assert loo.accuracy > 0.74
     assert loo.weighted_accuracy > loo.accuracy, "Weighted accuracy (#{loo.weighted_accuracy}) should be larger than accuracy (#{loo.accuracy})."
   end
 
-- 
cgit v1.2.3


From eef5d89a92dd7dde9acf9fc063a54e1fe729a89b Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Tue, 28 Mar 2017 11:00:12 +0000
Subject: fixed wrong accuracy assertion to rmse

---
 test/validation-regression.rb | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/validation-regression.rb b/test/validation-regression.rb
index 01ed644..c5ad312 100644
--- a/test/validation-regression.rb
+++ b/test/validation-regression.rb
@@ -84,7 +84,7 @@ class ValidationRegressionTest < MiniTest::Test
     repeated_cv = RepeatedCrossValidation.create model
     repeated_cv.crossvalidations.each do |cv|
       assert cv.r_squared > 0.34, "R^2 (#{cv.r_squared}) should be larger than 0.034"
-      assert_operator cv.accuracy, :>, 0.7, "model accuracy < 0.7, this may happen by chance due to an unfavorable training/test set split"
+      assert cv.rmse < 0.5, "RMSE (#{cv.rmse}) should be smaller than 0.5"
     end
   end
 
-- 
cgit v1.2.3


From db38c345fdc119edd8a892a5b0ba2c2a4b1cbe1f Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Fri, 31 Mar 2017 15:07:28 +0000
Subject: set default min sim to 0.1 for classification and 0.5 for regression

---
 lib/model.rb                      | 12 ++++++++----
 test/model-classification.rb      | 22 +++++++++++-----------
 test/validation-classification.rb |  4 ++--
 3 files changed, 21 insertions(+), 17 deletions(-)

diff --git a/lib/model.rb b/lib/model.rb
index 7cc6765..80affd5 100644
--- a/lib/model.rb
+++ b/lib/model.rb
@@ -68,10 +68,6 @@ module OpenTox
               :method => "fingerprint",
               :type => "MP2D",
             },
-            :similarity => {
-              :method => "Algorithm::Similarity.tanimoto",
-              :min => 0.5,
-            },
             :feature_selection => nil
           }
 
@@ -79,10 +75,18 @@ module OpenTox
             model.algorithms[:prediction] = {
                 :method => "Algorithm::Classification.weighted_majority_vote",
             }
+            model.algorithms[:similarity] = {
+              :method => "Algorithm::Similarity.tanimoto",
+              :min => 0.1,
+            }
           elsif model.class == LazarRegression
             model.algorithms[:prediction] = {
               :method => "Algorithm::Caret.rf",
             }
+            model.algorithms[:similarity] = {
+              :method => "Algorithm::Similarity.tanimoto",
+              :min => 0.5,
+            }
           end
 
         elsif substance_classes.first == "OpenTox::Nanoparticle"
diff --git a/test/model-classification.rb b/test/model-classification.rb
index 0bb3e83..1424f6a 100644
--- a/test/model-classification.rb
+++ b/test/model-classification.rb
@@ -10,35 +10,35 @@ class LazarClassificationTest < MiniTest::Test
       },
       :similarity => {
         :method => "Algorithm::Similarity.tanimoto",
-        :min => 0.5
+        :min => 0.1
       },
-      :feature_selection => nil,
       :prediction => {
         :method => "Algorithm::Classification.weighted_majority_vote",
       },
+      :feature_selection => nil,
     }
     training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"hamster_carcinogenicity.csv")
     model = Model::Lazar.create  training_dataset: training_dataset
     assert_kind_of Model::LazarClassification, model
     assert_equal algorithms, model.algorithms
-    substance = training_dataset.substances[49]
+    substance = training_dataset.substances[10]
     prediction = model.predict substance
     assert_equal "false", prediction[:value]
     [ {
-      :compound => OpenTox::Compound.from_inchi("InChI=1S/C6H14N2O4/c1-5(10)2-8(7-12)3-6(11)4-9/h5-6,9-11H,2-4H2,1H3"),
+      :compound => OpenTox::Compound.from_inchi("InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"),
       :prediction => "false",
     },{
-      :compound => OpenTox::Compound.from_smiles("OCC(CN(CC(O)C)N=O)O"),
+      :compound => OpenTox::Compound.from_smiles("c1ccccc1NN"),
       :prediction => "false",
     } ].each do |example|
       prediction = model.predict example[:compound]
       assert_equal example[:prediction], prediction[:value]
     end
 
-    compound = Compound.from_smiles "O=NN1CCC1"
+    compound = Compound.from_smiles "CCO"
     prediction = model.predict compound
     assert_equal "true", prediction[:value]
-    #assert_equal ["false"], prediction[:measurements]
+    assert_equal ["false"], prediction[:measurements]
 
     # make a dataset prediction
     compound_dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.mini_log10.csv")
@@ -46,12 +46,12 @@ class LazarClassificationTest < MiniTest::Test
     assert_equal compound_dataset.compounds, prediction_dataset.compounds
 
     cid = prediction_dataset.compounds[7].id.to_s
-    assert_equal "Could not find similar substances with experimental data in the training dataset.", prediction_dataset.predictions[cid][:warnings][0]
+    assert_equal "Could not find similar substances with experimental data in the training dataset.", prediction_dataset.predictions[cid][:warning]
     prediction_dataset.predictions.each do |cid,pred|
-      assert_equal "Could not find similar substances with experimental data in the training dataset.", pred[:warnings][0] if pred[:value].nil?
+      assert_equal "Could not find similar substances with experimental data in the training dataset.", pred[:warning] if pred[:value].nil?
     end
     cid = Compound.from_smiles("CCOC(=O)N").id.to_s
-    assert_match "excluded", prediction_dataset.predictions[cid][:info]
+    assert_match "excluded", prediction_dataset.predictions[cid][:warning]
     # cleanup
     [training_dataset,model,compound_dataset,prediction_dataset].each{|o| o.delete}
   end
@@ -85,7 +85,7 @@ class LazarClassificationTest < MiniTest::Test
     model = Model::Lazar.create training_dataset: training_dataset
     t = Time.now
     2.times do
-      compound = Compound.from_smiles("OCC(CN(CC(O)C)N=O)O")
+      compound = Compound.from_smiles("Clc1ccccc1NN")
       prediction = model.predict compound
       assert_equal "1", prediction[:value]
     end
diff --git a/test/validation-classification.rb b/test/validation-classification.rb
index ac25b29..fb4c3e7 100644
--- a/test/validation-classification.rb
+++ b/test/validation-classification.rb
@@ -47,9 +47,9 @@ class ValidationClassificationTest < MiniTest::Test
     dataset = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
     model = Model::Lazar.create training_dataset: dataset
     loo = ClassificationLeaveOneOut.create model
-    assert_equal 77, loo.nr_unpredicted
+    assert_equal 14, loo.nr_unpredicted
     refute_empty loo.confusion_matrix
-    assert loo.accuracy > 0.74
+    assert loo.accuracy > 0.77
     assert loo.weighted_accuracy > loo.accuracy, "Weighted accuracy (#{loo.weighted_accuracy}) should be larger than accuracy (#{loo.accuracy})."
   end
 
-- 
cgit v1.2.3


From 6201c1f3814628499e168bd99fddc5b65eb32fb5 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Fri, 31 Mar 2017 15:08:14 +0000
Subject: adjusted classification tests for min sim 0.1

---
 test/model-classification.rb      | 8 +++++---
 test/validation-classification.rb | 2 +-
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/test/model-classification.rb b/test/model-classification.rb
index 1424f6a..f75598b 100644
--- a/test/model-classification.rb
+++ b/test/model-classification.rb
@@ -46,12 +46,14 @@ class LazarClassificationTest < MiniTest::Test
     assert_equal compound_dataset.compounds, prediction_dataset.compounds
 
     cid = prediction_dataset.compounds[7].id.to_s
-    assert_equal "Could not find similar substances with experimental data in the training dataset.", prediction_dataset.predictions[cid][:warning]
+    assert_equal "Could not find similar substances with experimental data in the training dataset.", prediction_dataset.predictions[cid][:warnings][0]
+    expectations = ["Cannot create prediction: Only one similar compound in the training set.",
+    "Could not find similar substances with experimental data in the training dataset."]
     prediction_dataset.predictions.each do |cid,pred|
-      assert_equal "Could not find similar substances with experimental data in the training dataset.", pred[:warning] if pred[:value].nil?
+      assert_includes expectations, pred[:warnings][0] if pred[:value].nil?
     end
     cid = Compound.from_smiles("CCOC(=O)N").id.to_s
-    assert_match "excluded", prediction_dataset.predictions[cid][:warning]
+    assert_match "excluded", prediction_dataset.predictions[cid][:info]
     # cleanup
     [training_dataset,model,compound_dataset,prediction_dataset].each{|o| o.delete}
   end
diff --git a/test/validation-classification.rb b/test/validation-classification.rb
index fb4c3e7..ce06063 100644
--- a/test/validation-classification.rb
+++ b/test/validation-classification.rb
@@ -47,7 +47,7 @@ class ValidationClassificationTest < MiniTest::Test
     dataset = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
     model = Model::Lazar.create training_dataset: dataset
     loo = ClassificationLeaveOneOut.create model
-    assert_equal 14, loo.nr_unpredicted
+    assert_equal 24, loo.nr_unpredicted
     refute_empty loo.confusion_matrix
     assert loo.accuracy > 0.77
     assert loo.weighted_accuracy > loo.accuracy, "Weighted accuracy (#{loo.weighted_accuracy}) should be larger than accuracy (#{loo.accuracy})."
-- 
cgit v1.2.3


From 324e502a3a8415512336b93fa4faf591d7904ad8 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Tue, 11 Apr 2017 08:03:15 +0000
Subject: changed compound to predict fixed test

---
 test/model-validation.rb | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/model-validation.rb b/test/model-validation.rb
index 83986d6..9304232 100644
--- a/test/model-validation.rb
+++ b/test/model-validation.rb
@@ -12,7 +12,7 @@ class ValidationModelTest < MiniTest::Test
     m.crossvalidations.each do |cv|
       assert cv.accuracy > 0.74, "Crossvalidation accuracy (#{cv.accuracy}) should be larger than 0.75. This may happen due to an unfavorable training/test set split."
     end
-    prediction = m.predict Compound.from_smiles("CCCC(NN)C")
+    prediction = m.predict Compound.from_smiles("OCC(CN(CC(O)C)N=O)O")
     assert_equal "true", prediction[:value]
     m.delete
   end
-- 
cgit v1.2.3


From 1f95bfdc11de7c4fdc2e9f0313be3506dd6ea9c1 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Thu, 4 May 2017 09:50:08 +0000
Subject: ensure warnings for validation statistics

---
 lib/crossvalidation.rb       | 1 +
 lib/validation-statistics.rb | 8 ++++++--
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/lib/crossvalidation.rb b/lib/crossvalidation.rb
index 75c5db5..06a1e2a 100644
--- a/lib/crossvalidation.rb
+++ b/lib/crossvalidation.rb
@@ -90,6 +90,7 @@ module OpenTox
       field :within_prediction_interval, type: Integer, default:0
       field :out_of_prediction_interval, type: Integer, default:0
       field :correlation_plot_id, type: BSON::ObjectId
+      field :warnings, type: Array
     end
 
     # Independent repeated crossvalidations
diff --git a/lib/validation-statistics.rb b/lib/validation-statistics.rb
index 2d522ae..69e7992 100644
--- a/lib/validation-statistics.rb
+++ b/lib/validation-statistics.rb
@@ -111,6 +111,7 @@ module OpenTox
       # Get statistics
       # @return [Hash]
       def statistics
+        self.warnings = []
         self.rmse = 0
         self.mae = 0
         self.within_prediction_interval = 0
@@ -132,8 +133,10 @@ module OpenTox
               end
             end
           else
-            warnings << "No training activities for #{Compound.find(compound_id).smiles} in training dataset #{model.training_dataset_id}."
-            $logger.debug "No training activities for #{Compound.find(compound_id).smiles} in training dataset #{model.training_dataset_id}."
+            trd_id = model.training_dataset_id
+            smiles = Compound.find(cid).smiles
+            self.warnings << "No training activities for #{smiles} in training dataset #{trd_id}."
+            $logger.debug "No training activities for #{smiles} in training dataset #{trd_id}."
           end
         end
         R.assign "measurement", x
@@ -146,6 +149,7 @@ module OpenTox
         $logger.debug "RMSE #{rmse}"
         $logger.debug "MAE #{mae}"
         $logger.debug "#{percent_within_prediction_interval.round(2)}% of measurements within prediction interval"
+        $logger.debug "#{warnings}"
         save
         {
           :mae => mae,
-- 
cgit v1.2.3


From d4d914052de0a37489f9491dbe491093bd14a03a Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Thu, 4 May 2017 09:50:57 +0000
Subject: ensure always two probability values for classification

---
 lib/classification.rb | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/lib/classification.rb b/lib/classification.rb
index 638492b..a875903 100644
--- a/lib/classification.rb
+++ b/lib/classification.rb
@@ -18,6 +18,11 @@ module OpenTox
         class_weights.each do |a,w|
           probabilities[a] = w.sum/weights.sum
         end
+        # DG: hack to ensure always two probability values
+        if probabilities.keys.uniq.size == 1
+          missing_key = probabilities.keys.uniq[0].match(/^non/) ? probabilities.keys.uniq[0].sub(/non-/,"") : "non-"+probabilities.keys.uniq[0]
+          probabilities[missing_key] = 0.0
+        end
         probabilities = probabilities.collect{|a,p| [a,weights.max*p]}.to_h
         p_max = probabilities.collect{|a,p| p}.max
         prediction = probabilities.key(p_max)
-- 
cgit v1.2.3


From 658e0f706622eabce6900134fb1d968b440fd704 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Thu, 4 May 2017 09:59:37 +0000
Subject: cleanup and hack for mongos file size limit

---
 lib/model.rb                 | 1 -
 lib/train-test-validation.rb | 2 ++
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/lib/model.rb b/lib/model.rb
index 80affd5..9c5c19b 100644
--- a/lib/model.rb
+++ b/lib/model.rb
@@ -145,7 +145,6 @@ module OpenTox
           end
           model.descriptor_ids = model.fingerprints.flatten.uniq
           model.descriptor_ids.each do |d|
-            # resulting model may break BSON size limit (e.g. f Kazius dataset)
             model.independent_variables << model.substance_ids.collect_with_index{|s,i| model.fingerprints[i].include? d} if model.algorithms[:prediction][:method].match /Caret/
           end
         # calculate physchem properties
diff --git a/lib/train-test-validation.rb b/lib/train-test-validation.rb
index 034ae3a..9a5532d 100644
--- a/lib/train-test-validation.rb
+++ b/lib/train-test-validation.rb
@@ -27,6 +27,8 @@ module OpenTox
           end
         end
         predictions.select!{|cid,p| p[:value] and p[:measurements]}
+        # hack to avoid mongos file size limit error on large datasets
+        #predictions.each{|cid,p| p[:neighbors] = []} if model.training_dataset.name.match(/mutagenicity/i)
         validation = self.new(
           :model_id => validation_model.id,
           :test_dataset_id => test_set.id,
-- 
cgit v1.2.3


From f82f516db24cd688f8a75b2b9a27c5ac46aade67 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Fri, 5 May 2017 12:43:01 +0000
Subject: added warnings field for leave-one-out test fails

---
 lib/leave-one-out-validation.rb | 1 +
 1 file changed, 1 insertion(+)

diff --git a/lib/leave-one-out-validation.rb b/lib/leave-one-out-validation.rb
index 8d22018..c33c92b 100644
--- a/lib/leave-one-out-validation.rb
+++ b/lib/leave-one-out-validation.rb
@@ -58,6 +58,7 @@ module OpenTox
       field :within_prediction_interval, type: Integer, default:0
       field :out_of_prediction_interval, type: Integer, default:0
       field :correlation_plot_id, type: BSON::ObjectId
+      field :warnings, type: Array
     end
 
   end
-- 
cgit v1.2.3


From 9aa5203dd375225996c1efe4be1a4324ddc6cda7 Mon Sep 17 00:00:00 2001
From: gebele <gebele@in-silico.ch>
Date: Mon, 22 May 2017 12:45:32 +0000
Subject: fix for lower min sim threshold

---
 lib/model.rb | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/model.rb b/lib/model.rb
index 9c5c19b..475a346 100644
--- a/lib/model.rb
+++ b/lib/model.rb
@@ -276,7 +276,7 @@ module OpenTox
             #prediction[:warnings] << "Closest neighbor has similarity < #{algorithms[:similarity][:warn_min]}. Prediction may be out of applicability domain."
           #end
         end
-        if prediction[:warnings].empty? or threshold <= algorithms[:similarity][:min]
+        if prediction[:warnings].empty? or threshold < algorithms[:similarity][:min] or threshold <= 0.2
           prediction
         else # try again with a lower threshold
           predict_substance substance, 0.2
-- 
cgit v1.2.3