neighbor search delegated to database backend

10 files changed, 102 insertions, 29 deletions

diff --git a/test/compound.rb b/test/compound.rb
index 22c152b..ff20c1c 100644
--- a/test/compound.rb
+++ b/test/compound.rb
@@ -162,7 +162,7 @@ print c.sdf
   end
 
   def test_fingerprint_db_neighbors
-    skip
+    #skip
     training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.csv")
     [
       "CC(=O)CC(C)C#N",
@@ -170,8 +170,18 @@ print c.sdf
       "C(=O)CC(C)C#N",
     ].each do |smi|
       c = OpenTox::Compound.from_smiles smi
+      t = Time.now
       neighbors = c.db_neighbors(:training_dataset_id => training_dataset.id, :min_sim => 0.2)
-      p neighbors
+      p Time.now - t
+      t = Time.now
+      neighbors2 = c.fingerprint_neighbors({:type => "MP2D", :training_dataset_id => training_dataset.id, :min_sim => 0.2})
+      p Time.now - t
+      p neighbors.size
+      p neighbors2.size
+      #p neighbors
+      #p neighbors2
+      #p neighbors2 - neighbors
+      #assert_equal neighbors, neighbors2
     end
   end
 end
diff --git a/test/dataset-long.rb b/test/dataset-long.rb
index 5c8dfb8..49b61df 100644
--- a/test/dataset-long.rb
+++ b/test/dataset-long.rb
@@ -86,6 +86,7 @@ class DatasetLongTest < MiniTest::Test
   end
 
   def test_upload_feature_dataset
+    skip
     t = Time.now
     f = File.join DATA_DIR, "rat_feature_dataset.csv"
     d = Dataset.from_csv_file f
diff --git a/test/dataset.rb b/test/dataset.rb
index 4f1e885..1814081 100644
--- a/test/dataset.rb
+++ b/test/dataset.rb
@@ -127,7 +127,7 @@ class DatasetTest < MiniTest::Test
     original_csv.shift
     csv.each_with_index do |row,i|
       compound = Compound.from_smiles row.shift
-      original_compound = Compound.from_smiles original_csv[i].shift
+      original_compound = Compound.from_smiles original_csv[i].shift.strip
       assert_equal original_compound.inchi, compound.inchi
       row.each_with_index do |v,j|
         if v.numeric?
@@ -142,7 +142,6 @@ class DatasetTest < MiniTest::Test
 
   def test_from_csv
     d = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
-    p d
     assert_equal Dataset, d.class
     assert_equal 1, d.features.size
     assert_equal 85, d.compounds.size
@@ -170,8 +169,7 @@ class DatasetTest < MiniTest::Test
   def test_from_csv2
     File.open("#{DATA_DIR}/temp_test.csv", "w+") { |file| file.write("SMILES,Hamster\nCC=O,true\n ,true\nO=C(N),true") }
     dataset = Dataset.from_csv_file "#{DATA_DIR}/temp_test.csv"
-    p dataset.warnings
-    assert_equal "Cannot parse SMILES compound ' ' at position 3, all entries are ignored.",  dataset.warnings.join
+    assert_equal "Cannot parse SMILES compound '' at position 3, all entries are ignored.",  dataset.warnings.join
     File.delete "#{DATA_DIR}/temp_test.csv"
     dataset.features.each{|f| feature = Feature.find f.id; feature.delete}
     dataset.delete
diff --git a/test/fminer-long.rb b/test/fminer-long.rb
index 0f202b4..845ed71 100644
--- a/test/fminer-long.rb
+++ b/test/fminer-long.rb
@@ -3,6 +3,7 @@ require_relative "setup.rb"
 class FminerTest < MiniTest::Test
 
   def test_fminer_multicell
+    skip
     #skip "multicell segfaults"
     # TODO aborts, probably fminer
     # or OpenBabel segfault
@@ -15,6 +16,7 @@ class FminerTest < MiniTest::Test
   end
 
   def test_fminer_isscan
+    skip
     dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"ISSCAN-multi.csv")
     feature_dataset = OpenTox::Algorithm::Fminer.bbrc(dataset)#, :min_frequency => 15)
     assert_equal feature_dataset.compounds.size, dataset.compounds.size
@@ -25,6 +27,7 @@ class FminerTest < MiniTest::Test
   end
 
   def test_fminer_kazius
+    skip
     dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"kazius.csv")
     # TODO reactivate default settings
     feature_dataset = OpenTox::Algorithm::Fminer.bbrc(dataset, :min_frequency => 20)
diff --git a/test/lazar-classification.rb b/test/lazar-classification.rb
new file mode 100644
index 0000000..e8b2181
--- /dev/null
+++ b/test/lazar-classification.rb
@@ -0,0 +1,42 @@
+require_relative "setup.rb"
+
+class LazarClassificationTest < MiniTest::Test
+
+  def test_lazar_classification
+    training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"hamster_carcinogenicity.csv")
+    model = Model::LazarClassification.create training_dataset#, feature_dataset
+    #assert_equal 'C-C-C=C', feature_dataset.features.first.smarts
+
+    [ {
+      :compound => OpenTox::Compound.from_inchi("InChI=1S/C6H6/c1-2-4-6-5-3-1/h1-6H"),
+      :prediction => "false",
+      :confidence => 0.25281385281385277,
+      :nr_neighbors => 11
+    },{
+      :compound => OpenTox::Compound.from_smiles("c1ccccc1NN"),
+      :prediction => "false",
+      :confidence => 0.3639589577089577,
+      :nr_neighbors => 14
+    } ].each do |example|
+      prediction = model.predict example[:compound]
+      assert_equal example[:prediction], prediction[:value]
+      #assert_equal example[:confidence], prediction[:confidence]
+      #assert_equal example[:nr_neighbors], prediction[:neighbors].size
+    end
+
+    compound = Compound.from_smiles "CCO"
+    prediction = model.predict compound
+    assert_equal ["false"], prediction[:database_activities]
+    assert_equal "true", prediction[:value]
+
+    # make a dataset prediction
+    compound_dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"EPAFHM.mini.csv")
+    prediction = model.predict compound_dataset
+    assert_equal compound_dataset.compounds, prediction.compounds
+
+    assert_equal "Cound not find similar compounds.", prediction.data_entries[7][2]
+    assert_equal "measured", prediction.data_entries[14][1]
+    # cleanup
+    [training_dataset,model,compound_dataset].each{|o| o.delete}
+  end
+end
diff --git a/test/lazar-fminer.rb b/test/lazar-fminer.rb
index 41e1071..9e024a1 100644
--- a/test/lazar-fminer.rb
+++ b/test/lazar-fminer.rb
@@ -3,6 +3,7 @@ require_relative "setup.rb"
 class LazarFminerTest < MiniTest::Test
 
   def test_lazar_fminer
+    skip
     training_dataset = Dataset.from_csv_file File.join(DATA_DIR,"hamster_carcinogenicity.csv")
     model = Model::LazarFminerClassification.create training_dataset#, feature_dataset
     feature_dataset = Dataset.find model.neighbor_algorithm_parameters[:feature_dataset_id]
diff --git a/test/lazar-long.rb b/test/lazar-long.rb
index 92d7d5a..525b96e 100644
--- a/test/lazar-long.rb
+++ b/test/lazar-long.rb
@@ -3,6 +3,7 @@ require_relative "setup.rb"
 class LazarExtendedTest < MiniTest::Test
 
   def test_lazar_bbrc_ham_minfreq
+    skip
     dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"hamster_carcinogenicity.csv")
     model = Model::LazarFminerClassification.create(dataset, :min_frequency => 5)
     feature_dataset = Dataset.find model.neighbor_algorithm_parameters[:feature_dataset_id]
@@ -21,6 +22,7 @@ class LazarExtendedTest < MiniTest::Test
   end
 
   def test_lazar_bbrc_large_ds
+    skip
     dataset = OpenTox::Dataset.from_csv_file File.join(DATA_DIR,"multi_cell_call_no_dup.csv")
     model = Model::LazarFminerClassification.create dataset
     feature_dataset = Dataset.find model.neighbor_algorithm_parameters[:feature_dataset_id]
@@ -44,7 +46,8 @@ class LazarExtendedTest < MiniTest::Test
     feature_dataset.delete
   end
 
-  def test_lazar_kazius
+  def test_lazar_fminer_kazius
+    skip
     t = Time.now
     dataset = Dataset.from_csv_file File.join(DATA_DIR,"kazius.csv")
     p "Dataset upload: #{Time.now-t}"
@@ -68,4 +71,22 @@ class LazarExtendedTest < MiniTest::Test
     #feature_dataset.delete
   end
 
+  def test_lazar_kazius
+    t = Time.now
+    dataset = Dataset.from_csv_file File.join(DATA_DIR,"kazius.csv")
+    p "Dataset upload: #{Time.now-t}"
+    t = Time.now
+    model = Model::LazarClassification.create(dataset)
+    p "Feature mining: #{Time.now-t}"
+    t = Time.now
+    2.times do
+      compound = Compound.from_smiles("Clc1ccccc1NN")
+      prediction = model.predict compound
+      #p prediction
+      assert_equal "1", prediction[:value]
+      #assert_in_delta 0.019858401199860445, prediction[:confidence], 0.001
+    end
+    dataset.delete
+  end
+
 end
diff --git a/test/lazar-regression.rb b/test/lazar-regression.rb
index 4f5a332..c1dc9b9 100644
--- a/test/lazar-regression.rb
+++ b/test/lazar-regression.rb
@@ -8,7 +8,7 @@ class LazarRegressionTest < MiniTest::Test
     compound = Compound.from_smiles "CC(C)(C)CN"
     prediction = model.predict compound
     assert_equal 7.2, prediction[:value].round(1)
-    assert_equal 91, prediction[:neighbors].size
+    assert_equal 88, prediction[:neighbors].size
   end
 
   def test_mpd_fingerprints
@@ -17,7 +17,7 @@ class LazarRegressionTest < MiniTest::Test
     model.neighbor_algorithm_parameters[:type] = "MP2D"
     compound = Compound.from_smiles "CCCSCCSCC"
     prediction = model.predict compound
-    assert_equal 0.02, prediction[:value].round(2)
+    assert_equal 0.04, prediction[:value].round(2)
     assert_equal 3, prediction[:neighbors].size
   end
 
diff --git a/test/prediction_models.rb b/test/prediction_models.rb
index 1b9e788..067c3c8 100644
--- a/test/prediction_models.rb
+++ b/test/prediction_models.rb
@@ -4,22 +4,13 @@ class PredictionModelTest < MiniTest::Test
 
   def test_prediction_model
     pm = Model::Prediction.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
-    #dataset = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
-    #model = Model::LazarFminerClassification.create dataset
-    #cv = ClassificationCrossValidation.create model
-    #metadata = JSON.parse(File.read("#{DATA_DIR}/hamster_carcinogenicity.json"))
-
-    #metadata[:model_id] = model.id
-    #metadata[:crossvalidation_id] = cv.id
-    #pm = Model::Prediction.new(metadata)
-    #pm.save
     [:endpoint,:species,:source].each do |p|
       refute_empty pm[p]
     end
     assert pm.classification?
     refute pm.regression?
     pm.crossvalidations.each do |cv|
-      assert cv.accuracy > 0.75
+      assert cv.accuracy > 0.75, "Crossvalidation accuracy (#{cv.accuracy}) should be larger than 0.75. This may happen due to an unfavorable training/test set split."
     end
     prediction = pm.predict Compound.from_smiles("CCCC(NN)C")
     assert_equal "true", prediction[:value]
diff --git a/test/validation.rb b/test/validation.rb
index 6764a32..7de944c 100644
--- a/test/validation.rb
+++ b/test/validation.rb
@@ -3,6 +3,7 @@ require_relative "setup.rb"
 class ValidationTest < MiniTest::Test
 
   def test_fminer_crossvalidation
+    skip
     dataset = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
     model = Model::LazarFminerClassification.create dataset
     cv = ClassificationCrossValidation.create model
@@ -15,12 +16,13 @@ class ValidationTest < MiniTest::Test
     dataset = Dataset.from_csv_file "#{DATA_DIR}/hamster_carcinogenicity.csv"
     model = Model::LazarClassification.create dataset#, features
     cv = ClassificationCrossValidation.create model
-    assert cv.accuracy > 0.7
-    File.open("tmp.svg","w+"){|f| f.puts cv.confidence_plot}
-    `inkview tmp.svg`
+    #p cv
+    assert cv.accuracy > 0.7, "Accuracy (#{cv.accuracy}) should be larger than 0.7"
+    #File.open("tmp.svg","w+"){|f| f.puts cv.confidence_plot}
+    #`inkview tmp.svg`
     p cv.nr_unpredicted
     p cv.accuracy
-    #assert cv.weighted_accuracy > cv.accuracy, "Weighted accuracy should be larger than unweighted accuracy."
+    assert cv.weighted_accuracy > cv.accuracy, "Weighted accuracy (#{cv.weighted_accuracy}) should be larger than unweighted accuracy (#{cv.accuracy}) ."
   end
 
   def test_default_regression_crossvalidation
@@ -28,11 +30,11 @@ class ValidationTest < MiniTest::Test
     model = Model::LazarRegression.create dataset
     cv = RegressionCrossValidation.create model
     #cv = RegressionCrossValidation.find '561503262b72ed54fd000001'
-    p cv.id
-    File.open("tmp.svg","w+"){|f| f.puts cv.correlation_plot}
-    `inkview tmp.svg`
-    File.open("tmp.svg","w+"){|f| f.puts cv.confidence_plot}
-    `inkview tmp.svg`
+    #p cv.id
+    #File.open("tmp.svg","w+"){|f| f.puts cv.correlation_plot}
+    #`inkview tmp.svg`
+    #File.open("tmp.svg","w+"){|f| f.puts cv.confidence_plot}
+    #`inkview tmp.svg`
     
     #puts cv.misclassifications.to_yaml
     p cv.rmse
@@ -91,9 +93,13 @@ class ValidationTest < MiniTest::Test
     model.save
     cv = ClassificationCrossValidation.create model
     params = model.neighbor_algorithm_parameters
+    params.delete :training_dataset_id
     params = Hash[params.map{ |k, v| [k.to_s, v] }] # convert symbols to string
+
     cv.validations.each do |validation|
-      assert_equal params, validation.model.neighbor_algorithm_parameters
+      validation_params = validation.model.neighbor_algorithm_parameters
+      validation_params.delete "training_dataset_id"
+      assert_equal params, validation_params
     end
   end