From 719b7fd84a1593fe209c393d5858fce64c7314aa Mon Sep 17 00:00:00 2001 From: Christoph Helma Date: Thu, 15 Mar 2018 16:05:55 +0100 Subject: warnings improved --- 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