path: root/fminer.rb

ENV['FMINER_SMARTS'] = 'true'
ENV['FMINER_NO_AROMATIC'] = 'true'
ENV['FMINER_PVALUES'] = 'true'
ENV['FMINER_SILENT'] = 'true'
ENV['FMINER_NR_HITS'] = 'true'

@@bbrc = Bbrc::Bbrc.new
@@last = Last::Last.new



# Get list of fminer algorithms
#
# @return [text/uri-list] URIs of fminer algorithms
get '/fminer/?' do
  list = [ url_for('/fminer/bbrc', :full), url_for('/fminer/bbrc/sample', :full), url_for('/fminer/last', :full), url_for('/fminer/bbrc/match', :full), url_for('/fminer/last/match', :full) ].join("\n") + "\n"
  case request.env['HTTP_ACCEPT']
  when /text\/html/
    content_type "text/html"
    OpenTox.text_to_html list
  else
    content_type 'text/uri-list'
    list
  end
end



# Get RDF/XML representation of fminer bbrc algorithm
# @return [application/rdf+xml] OWL-DL representation of fminer bbrc algorithm
get "/fminer/bbrc/?" do
  algorithm = OpenTox::Algorithm::Generic.new(url_for('/fminer/bbrc',:full))
  algorithm.metadata = {
    DC.title => 'fminer backbone refinement class representatives',
    DC.creator => "andreas@maunz.de, helma@in-silico.ch",
    DC.contributor => "vorgrimmlerdavid@gmx.de",
#    BO.instanceOf => "http://opentox.org/ontology/ist-algorithms.owl#fminer_bbrc",
    RDF.type => [OT.Algorithm,OTA.PatternMiningSupervised],
    OT.parameters => [
      { DC.description => "Dataset URI", OT.paramScope => "mandatory", DC.title => "dataset_uri" },
      { DC.description => "Feature URI for dependent variable", OT.paramScope => "mandatory", DC.title => "prediction_feature" },
      { DC.description => "Minimum frequency", OT.paramScope => "optional", DC.title => "min_frequency" },
      { DC.description => "Feature type, can be 'paths' or 'trees'", OT.paramScope => "optional", DC.title => "feature_type" },
      { DC.description => "BBRC classes, pass 'false' to switch off mining for BBRC representatives.", OT.paramScope => "optional", DC.title => "backbone" },
      { DC.description => "Significance threshold (between 0 and 1)", OT.paramScope => "optional", DC.title => "min_chisq_significance" },
      { DC.description => "Whether subgraphs should be weighted with their occurrence counts in the instances (frequency)", OT.paramScope => "optional", DC.title => "nr_hits" },
  ]
  }
  case request.env['HTTP_ACCEPT']
  when /text\/html/
    content_type "text/html"
    OpenTox.text_to_html algorithm.to_yaml
  when /application\/x-yaml/
    content_type "application/x-yaml"
    algorithm.to_yaml
  else
    response['Content-Type'] = 'application/rdf+xml'
    algorithm.to_rdfxml
  end
end

# Get RDF/XML representation of fminer bbrc algorithm
# @return [application/rdf+xml] OWL-DL representation of fminer bbrc algorithm
get "/fminer/bbrc/sample/?" do
  algorithm = OpenTox::Algorithm::Generic.new(url_for('/fminer/bbrc/sample',:full))
  algorithm.metadata = {
    DC.title => 'fminer backbone refinement class representatives, obtained from samples of a dataset',
    DC.creator => "andreas@maunz.de",
#    BO.instanceOf => "http://opentox.org/ontology/ist-algorithms.owl#fminer_bbrc",
    RDF.type => [OT.Algorithm,OTA.PatternMiningSupervised],
    OT.parameters => [
      { DC.description => "Dataset URI", OT.paramScope => "mandatory", DC.title => "dataset_uri" },
      { DC.description => "Feature URI for dependent variable", OT.paramScope => "mandatory", DC.title => "prediction_feature" },
      { DC.description => "Number of bootstrap samples", OT.paramScope => "optional", DC.title => "num_boots" },
      { DC.description => "Minimum sampling support", OT.paramScope => "optional", DC.title => "min_sampling_support" },
      { DC.description => "Minimum frequency", OT.paramScope => "optional", DC.title => "min_frequency" },
      { DC.description => "Whether subgraphs should be weighted with their occurrence counts in the instances (frequency)", OT.paramScope => "optional", DC.title => "nr_hits" },
      { DC.description => "BBRC classes, pass 'false' to switch off mining for BBRC representatives.", OT.paramScope => "optional", DC.title => "backbone" },
      { DC.description => "Chisq estimation method, pass 'mean' to use simple mean estimate for chisq test.", OT.paramScope => "optional", DC.title => "method" }
  ]
  }
  case request.env['HTTP_ACCEPT']
  when /text\/html/
    content_type "text/html"
    OpenTox.text_to_html algorithm.to_yaml
  when /yaml/
    content_type "application/x-yaml"
    algorithm.to_yaml
  else
    response['Content-Type'] = 'application/rdf+xml'
    algorithm.to_rdfxml
  end
end

# Get RDF/XML representation of fminer last algorithm
# @return [application/rdf+xml] OWL-DL representation of fminer last algorithm
get "/fminer/last/?" do
  algorithm = OpenTox::Algorithm::Generic.new(url_for('/fminer/last',:full))
  algorithm.metadata = {
    DC.title => 'fminer latent structure class representatives',
    DC.creator => "andreas@maunz.de, helma@in-silico.ch",
    DC.contributor => "vorgrimmlerdavid@gmx.de",
#    BO.instanceOf => "http://opentox.org/ontology/ist-algorithms.owl#fminer_last",
    RDF.type => [OT.Algorithm,OTA.PatternMiningSupervised],
    OT.parameters => [
      { DC.description => "Dataset URI", OT.paramScope => "mandatory", DC.title => "dataset_uri" },
      { DC.description => "Feature URI for dependent variable", OT.paramScope => "mandatory", DC.title => "prediction_feature" },
      { DC.description => "Minimum frequency", OT.paramScope => "optional", DC.title => "min_frequency" },
      { DC.description => "Feature type, can be 'paths' or 'trees'", OT.paramScope => "optional", DC.title => "feature_type" },
      { DC.description => "Whether subgraphs should be weighted with their occurrence counts in the instances (frequency)", OT.paramScope => "optional", DC.title => "nr_hits" },
  ]
  }
  case request.env['HTTP_ACCEPT']
  when /text\/html/
    content_type "text/html"
    OpenTox.text_to_html algorithm.to_yaml
  when /application\/x-yaml/
    content_type "application/x-yaml"
    algorithm.to_yaml
  else
    response['Content-Type'] = 'application/rdf+xml'
    algorithm.to_rdfxml
  end
end


# Get RDF/XML representation of fminer matching algorithm
# @param [String] dataset_uri URI of the dataset 
# @param [String] feature_dataset_uri URI of the feature dataset (i.e. dependent variable)
# @param [optional] parameters Accepted parameters are
# - prediction_feature URI of prediction feature to calculate p-values for
get "/fminer/:method/match?" do
  algorithm = OpenTox::Algorithm::Generic.new(url_for("/fminer/#{params[:method]}/match",:full))
  algorithm.metadata = {
    DC.title => 'fminer feature matching',
    DC.creator => "mguetlein@gmail.com, andreas@maunz.de",
    RDF.type => [OT.Algorithm,OTA.PatternMiningSupervised],
    OT.parameters => [
      { DC.description => "Dataset URI", OT.paramScope => "mandatory", DC.title => "dataset_uri" },
      { DC.description => "Feature Dataset URI", OT.paramScope => "mandatory", DC.title => "feature_dataset_uri" },
      { DC.description => "Feature URI for dependent variable", OT.paramScope => "optional", DC.title => "prediction_feature" }
  ]
  }
  case request.env['HTTP_ACCEPT']
  when /text\/html/
    content_type "text/html"
    OpenTox.text_to_html algorithm.to_yaml
  when /application\/x-yaml/
    content_type "application/x-yaml"
    algorithm.to_yaml
  else
    response['Content-Type'] = 'application/rdf+xml'
    algorithm.to_rdfxml
  end
end




# Run bbrc algorithm on dataset
#
# @param [String] dataset_uri URI of the training dataset
# @param [String] prediction_feature URI of the prediction feature (i.e. dependent variable)
# @param [optional] parameters BBRC parameters, accepted parameters are
#   - min_frequency  Minimum frequency (default 5)
#   - feature_type Feature type, can be 'paths' or 'trees' (default "trees")
#   - backbone BBRC classes, pass 'false' to switch off mining for BBRC representatives. (default "true")
#   - min_chisq_significance Significance threshold (between 0 and 1)
#   - nr_hits Set to "true" to get hit count instead of presence
# @return [text/uri-list] Task URI
post '/fminer/bbrc/?' do

  fminer=OpenTox::Algorithm::Fminer.new
  fminer.check_params(params,5,@subjectid)

  task = OpenTox::Task.create("Mining BBRC features", url_for('/fminer',:full)) do |task|
    @@bbrc.Reset
    if fminer.prediction_feature.feature_type == "regression"
      @@bbrc.SetRegression(true) # AM: DO NOT MOVE DOWN! Must happen before the other Set... operations!
    else
      raise "no accept values for dataset '"+fminer.training_dataset.uri.to_s+"' and feature '"+fminer.prediction_feature.uri.to_s+
        "'" unless fminer.training_dataset.accept_values(fminer.prediction_feature.uri)
      @value_map=fminer.training_dataset.value_map(fminer.prediction_feature.uri)
    end
    @@bbrc.SetMinfreq(fminer.minfreq)
    @@bbrc.SetType(1) if params[:feature_type] == "paths"
    @@bbrc.SetBackbone(false) if params[:backbone] == "false"
    @@bbrc.SetChisqSig(params[:min_chisq_significance].to_f) if params[:min_chisq_significance]
    @@bbrc.SetConsoleOut(false)

    feature_dataset = OpenTox::Dataset.new(nil, @subjectid)
    feature_dataset.add_metadata({
      DC.title => "BBRC representatives for " + fminer.training_dataset.metadata[DC.title].to_s,
      DC.creator => url_for('/fminer/bbrc',:full),
      OT.hasSource => url_for('/fminer/bbrc', :full),
      OT.parameters => [
        { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] },
        { DC.title => "prediction_feature", OT.paramValue => params[:prediction_feature] },
        { DC.title => "min_frequency", OT.paramValue => fminer.minfreq },
        { DC.title => "nr_hits", OT.paramValue => (params[:nr_hits] == "true" ? "true" : "false") },
        { DC.title => "backbone", OT.paramValue => (params[:backbone] == "false" ? "false" : "true") }

    ]
    })
    feature_dataset.save(@subjectid)

    fminer.compounds = []
    fminer.db_class_sizes = Array.new # AM: effect
    fminer.all_activities = Hash.new # DV: for effect calculation in regression part
    fminer.smi = [] # AM LAST: needed for matching the patterns back

    # Add data to fminer
    fminer.add_fminer_data(@@bbrc, @value_map)

    g_array=fminer.all_activities.values # DV: calculation of global median for effect calculation
    g_median=g_array.to_scale.median

    raise "No compounds in dataset #{fminer.training_dataset.uri}" if fminer.compounds.size==0
    task.progress 10
    step_width = 80 / @@bbrc.GetNoRootNodes().to_f
    features = Set.new

    # run @@bbrc
    (0 .. @@bbrc.GetNoRootNodes()-1).each do |j|
      results = @@bbrc.MineRoot(j)
      task.progress 10+step_width*(j+1)
      results.each do |result|
        f = YAML.load(result)[0]
        smarts = f[0]
        p_value = f[1]

        if (!@@bbrc.GetRegression)
          id_arrs = f[2..-1].flatten
          max = OpenTox::Algorithm.effect(f[2..-1].reverse, fminer.db_class_sizes) # f needs reversal for bbrc
          effect = max+1
        else #regression part
          id_arrs = f[2]
          # DV: effect calculation
          f_arr=Array.new
          f[2].each do |id|
            id=id.keys[0] # extract id from hit count hash
            f_arr.push(fminer.all_activities[id])
          end
          f_median=f_arr.to_scale.median
          if g_median >= f_median
            effect = 'activating'
          else
            effect = 'deactivating'
          end
        end

        feature_uri = File.join feature_dataset.uri,"feature","bbrc", features.size.to_s
        unless features.include? smarts
          features << smarts
          metadata = {
            OT.hasSource => url_for('/fminer/bbrc', :full),
            RDF.type => [OT.Feature, OT.Substructure],
            OT.smarts => smarts,
            OT.pValue => p_value.to_f,
            OT.effect => effect,
            OT.parameters => [
              { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] },
              { DC.title => "prediction_feature", OT.paramValue => params[:prediction_feature] }
          ]
          }
          feature_dataset.add_feature feature_uri, metadata
          #feature_dataset.add_feature_parameters feature_uri, feature_dataset.parameters
        end
        id_arrs.each { |id_count_hash|
          id=id_count_hash.keys[0].to_i
          count=id_count_hash.values[0].to_i
          if params[:nr_hits] == "true"
            feature_dataset.add(fminer.compounds[id], feature_uri, count)
          else
            feature_dataset.add(fminer.compounds[id], feature_uri, 1)
          end
        }

      end # end of
    end   # feature parsing

    # AM: add feature values for non-present features
    # feature_dataset.complete_data_entries

    feature_dataset.save(@subjectid)
    feature_dataset.uri
  end
  response['Content-Type'] = 'text/uri-list'
  raise OpenTox::ServiceUnavailableError.newtask.uri+"\n" if task.status == "Cancelled"
  halt 202,task.uri.to_s+"\n"
end
#end


# Run bbrc/sample algorithm on a dataset
#
# @param [String] dataset_uri URI of the training dataset
# @param [String] prediction_feature URI of the prediction feature (i.e. dependent variable)
# @param [optional] BBRC sample parameters, accepted are
#   - num_boots Number of bootstrap samples (default 150)
#   - min_sampling_support Minimum sampling support (default 30% of num_boots)
#   - min_frequency  Minimum frequency (default 10% of dataset size)
#   - nr_hits Whether subgraphs should be weighted with their occurrence counts in the instances (frequency)
#   - random_seed Random seed ensures same datasets in bootBbrc
#   - backbone BBRC classes, pass 'false' to switch off mining for BBRC representatives. (default "true")
#   - method Chisq estimation method, pass 'mean' to use simple mean estimate (default 'mle').
#   - cache Whether cache files should be used for the combination of dataset, min_frequency, backbone, random seed (default "false")
#
# @return [text/uri-list] Task URI
post '/fminer/bbrc/sample/?' do

  fminer=OpenTox::Algorithm::Fminer.new
  fminer.check_params(params,100,@subjectid) # AM: 100 per-mil (10%) as default minfreq

  # num_boots
  unless params[:num_boots]
    num_boots = 150
    LOGGER.debug "Set num_boots to default value #{num_boots}"
  else
    raise OpenTox::BadRequestError.new "num_boots is not numeric" unless OpenTox::Algorithm.numeric? params[:num_boots]
	  num_boots = params[:num_boots].to_i.ceil
  end

  # min_sampling_support
  unless params[:min_sampling_support]
    min_sampling_support = (num_boots * 0.3).ceil
    LOGGER.debug "Set min_sampling_support to default value #{min_sampling_support}"
  else
    raise OpenTox::BadRequestError.new "min_sampling_support is not numeric" unless OpenTox::Algorithm.numeric? params[:min_sampling_support]
	  min_sampling_support= params[:min_sampling_support].to_i.ceil
  end

  # random_seed
  unless params[:random_seed]
    random_seed = 1
    LOGGER.debug "Set random seed to default value #{random_seed}"
  else
    raise OpenTox::BadRequestError.new "random_seed is not numeric" unless OpenTox::Algorithm.numeric? params[:random_seed]
    random_seed= params[:random_seed].to_i.ceil
  end

  # backbone
  unless params[:backbone]
    backbone = "true"
    LOGGER.debug "Set backbone to default value #{backbone}"
  else
    raise OpenTox::BadRequestError.new "backbone is neither 'true' nor 'false'" unless (params[:backbone] == "true" or params[:backbone] == "false")
    backbone = params[:backbone]
  end

  # method
  unless params[:method]
    method="mle"
    LOGGER.debug "Set method to default value #{method}"
  else
    raise OpenTox::BadRequestError.new "method is neither 'mle' nor 'mean'" unless (params[:method] == "mle" or params[:method] == "mean")
    method = params[:method]
  end

  # cache
  cache=false
  if params[:cache] == "true"
    cache=true
    LOGGER.debug "Set cache to true"
  end


  task = OpenTox::Task.create("Mining BBRC sample features", url_for('/fminer',:full)) do |task|
    if fminer.prediction_feature.feature_type == "regression"
      raise OpenTox::BadRequestError.new "BBRC sampling is only for classification"
    else
      raise "no accept values for dataset '"+fminer.training_dataset.uri.to_s+"' and feature '"+fminer.prediction_feature.uri.to_s+
        "'" unless fminer.training_dataset.accept_values(fminer.prediction_feature.uri)
      @value_map=fminer.training_dataset.value_map(fminer.prediction_feature.uri)
    end

    feature_dataset = OpenTox::Dataset.new(nil, @subjectid)
    feature_dataset.add_metadata({
      DC.title => "BBRC representatives for " + fminer.training_dataset.metadata[DC.title].to_s + "(bootstrapped)",
      DC.creator => url_for('/fminer/bbrc/sample',:full),
      OT.hasSource => url_for('/fminer/bbrc/sample', :full)
    })
    feature_dataset.save(@subjectid)

    # filled by add_fminer_data:
    fminer.compounds = [] # indexed by id, starting from 1 (not 0)
    fminer.db_class_sizes = Array.new # for effect calculation
    fminer.all_activities = Hash.new # for effect calculation, indexed by id, starting from 1 (not 0)
    fminer.smi = [] # needed for matching the patterns back, indexed by id, starting from 1 (not 0)
    fminer.add_fminer_data(nil, @value_map) # To only fill in administrative data (no fminer priming) pass 'nil' as instance

    raise "No compounds in dataset #{fminer.training_dataset.uri}" if fminer.compounds.size==0


    # run bbrc-sample, obtain smarts and p-values
    features = Set.new
    task.progress 10
    @r = RinRuby.new(true,false) # global R instance leads to Socket errors after a large number of requests
    @r.assign "dataset.uri", params[:dataset_uri]
    @r.assign "prediction.feature.uri", fminer.prediction_feature.uri
    @r.assign "num.boots", num_boots
    @r.assign "min.frequency.per.sample", fminer.minfreq
    @r.assign "min.sampling.support", min_sampling_support
    @r.assign "random.seed", random_seed
    @r.assign "backbone", backbone
    @r.assign "bbrc.service", File.join(CONFIG[:services]["opentox-algorithm"], "fminer/bbrc")
    @r.assign "dataset.service", CONFIG[:services]["opentox-dataset"]
    @r.assign "method", method

    require 'digest/md5'
    fminer.smi.shift
    cachedId = Digest::MD5.hexdigest(
      fminer.smi.sort.join+
      num_boots.to_s+
      fminer.minfreq.to_s+
      random_seed.to_s+
      backbone.to_s
    )
    @r.assign "cachedId", cachedId
    @r.eval "cachedId <- NULL" unless cache

    @r.eval "source(\"bbrc-sample/bbrc-sample.R\")"
    begin
      @r.eval "bootBbrc(dataset.uri, prediction.feature.uri, num.boots, min.frequency.per.sample, min.sampling.support, cachedId, bbrc.service, dataset.service, T, random.seed, as.logical(backbone), method)"
      smarts = (@r.pull "ans.patterns").collect! { |id| id.gsub(/\'/,"") } # remove extra quotes around smarts
      r_p_values = @r.pull "ans.p.values"
      smarts_p_values = {}; smarts.size.times { |i| smarts_p_values[ smarts[i] ] = r_p_values[i] }
      merge_time = @r.pull "merge.time"
      n_stripped_mss = @r.pull "n.stripped.mss"
      n_stripped_cst = @r.pull "n.stripped.cst"
    rescue Exception => e
      LOGGER.debug "#{e.class}: #{e.message}"
      LOGGER.debug "Backtrace:\n\t#{e.backtrace.join("\n\t")}"
    end
    @r.quit # free R

    # matching
    task.progress 90
    lu = LU.new                             # AM LAST: uses last-utils here
    params[:nr_hits] == "true" ? hit_count=true: hit_count=false
    matches, counts = lu.match_rb(fminer.smi,smarts,hit_count)       # AM LAST: creates instantiations
    
    feature_dataset.add_metadata({
          OT.parameters => [
        { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] },
        { DC.title => "prediction_feature", OT.paramValue => params[:prediction_feature] },
        { DC.title => "min_sampling_support", OT.paramValue => min_sampling_support },
        { DC.title => "num_boots", OT.paramValue => num_boots },
        { DC.title => "min_frequency_per_sample", OT.paramValue => fminer.minfreq },
        { DC.title => "nr_hits", OT.paramValue => hit_count.to_s },
        { DC.title => "merge_time", OT.paramValue => merge_time.to_s },
        { DC.title => "n_stripped_mss", OT.paramValue => n_stripped_mss.to_s },
        { DC.title => "n_stripped_cst", OT.paramValue => n_stripped_cst.to_s },
        { DC.title => "random_seed", OT.paramValue => random_seed.to_s },
        { DC.title => "backbone", OT.paramValue => backbone.to_s },
        { DC.title => "method", OT.paramValue => method.to_s }
          ]
    })

    matches.each do |smarts, ids|
      feat_hash = Hash[*(fminer.all_activities.select { |k,v| ids.include?(k) }.flatten)] # AM LAST: get activities of feature occurrences; see http://www.softiesonrails.com/2007/9/18/ruby-201-weird-hash-syntax
      g = Array.new
      @value_map.each { |y,act| g[y-1]=Array.new }
      feat_hash.each  { |x,y|   g[y-1].push(x)   }
      max = OpenTox::Algorithm.effect(g, fminer.db_class_sizes)
      effect = max + 1
      feature_uri = File.join feature_dataset.uri,"feature","bbrc", features.size.to_s
      unless features.include? smarts
        features << smarts
        metadata = {
          RDF.type => [OT.Feature, OT.Substructure],
          OT.hasSource => feature_dataset.uri,
          OT.smarts => smarts,
          OT.pValue => smarts_p_values[smarts],
          OT.effect => effect,
          OT.parameters => [
            { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] },
            { DC.title => "prediction_feature", OT.paramValue => params[:prediction_feature] }
        ]
        }
        feature_dataset.add_feature feature_uri, metadata
      end
      if !hit_count
        ids.each { |id| feature_dataset.add(fminer.compounds[id], feature_uri, 1)}
      else
        ids.each_with_index { |id,i| feature_dataset.add(fminer.compounds[id], feature_uri, counts[smarts][i])}
      end
    end

    # AM: add feature values for non-present features
    # feature_dataset.complete_data_entries

    feature_dataset.save(@subjectid)
    feature_dataset.uri
  end
  response['Content-Type'] = 'text/uri-list'
  raise OpenTox::ServiceUnavailableError.newtask.uri+"\n" if task.status == "Cancelled"
  halt 202,task.uri.to_s+"\n"
end

# Run last algorithm on a dataset
#
# @param [String] dataset_uri URI of the training dataset
# @param [String] prediction_feature URI of the prediction feature (i.e. dependent variable)
# @param [optional] parameters LAST parameters, accepted parameters are
#   - min_frequency freq  Minimum frequency (default 5)
#   - feature_type Feature type, can be 'paths' or 'trees' (default "trees")
#   - nr_hits Set to "true" to get hit count instead of presence
# @return [text/uri-list] Task URI
post '/fminer/last/?' do

  fminer=OpenTox::Algorithm::Fminer.new
  fminer.check_params(params,80,@subjectid)

  task = OpenTox::Task.create("Mining LAST features", url_for('/fminer',:full)) do |task|
    @@last.Reset
    if fminer.prediction_feature.feature_type == "regression"
      @@last.SetRegression(true) # AM: DO NOT MOVE DOWN! Must happen before the other Set... operations!
    else
      raise "no accept values for dataset '"+fminer.training_dataset.uri.to_s+"' and feature '"+fminer.prediction_feature.uri.to_s+
        "'" unless fminer.training_dataset.accept_values(fminer.prediction_feature.uri)
      @value_map=fminer.training_dataset.value_map(fminer.prediction_feature.uri)
    end
    @@last.SetMinfreq(fminer.minfreq)
    @@last.SetType(1) if params[:feature_type] == "paths"
    @@last.SetConsoleOut(false)


    feature_dataset = OpenTox::Dataset.new(nil, @subjectid)
    feature_dataset.add_metadata({
      DC.title => "LAST representatives for " + fminer.training_dataset.metadata[DC.title].to_s,
      DC.creator => url_for('/fminer/last',:full),
      OT.hasSource => url_for('/fminer/last', :full),
      OT.parameters => [
        { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] },
        { DC.title => "prediction_feature", OT.paramValue => params[:prediction_feature] },
        { DC.title => "min_frequency", OT.paramValue => fminer.minfreq },
        { DC.title => "nr_hits", OT.paramValue => (params[:nr_hits] == "true" ? "true" : "false") }
      ]
    })
    feature_dataset.save(@subjectid)

    fminer.compounds = []
    fminer.db_class_sizes = Array.new # AM: effect
    fminer.all_activities = Hash.new # DV: for effect calculation (class and regr)
    fminer.smi = [] # AM LAST: needed for matching the patterns back

    # Add data to fminer
    fminer.add_fminer_data(@@last, @value_map)

    raise "No compounds in dataset #{fminer.training_dataset.uri}" if fminer.compounds.size==0

    # run @@last
    features = Set.new
    xml = ""
    task.progress 10
    step_width = 80 / @@last.GetNoRootNodes().to_f

    (0 .. @@last.GetNoRootNodes()-1).each do |j|
      results = @@last.MineRoot(j)
      task.progress 10+step_width*(j+1)
      results.each do |result|
        xml << result
      end
    end

    lu = LU.new                             # AM LAST: uses last-utils here
    dom=lu.read(xml)                        # AM LAST: parse GraphML
    smarts=lu.smarts_rb(dom,'nls')          # AM LAST: converts patterns to LAST-SMARTS using msa variant (see last-pm.maunz.de)
    params[:nr_hits] == "true" ? hit_count=true: hit_count=false
    matches, counts = lu.match_rb(fminer.smi,smarts,hit_count)       # AM LAST: creates instantiations

    matches.each do |smarts, ids|
      feat_hash = Hash[*(fminer.all_activities.select { |k,v| ids.include?(k) }.flatten)] # AM LAST: get activities of feature occurrences; see http://www.softiesonrails.com/2007/9/18/ruby-201-weird-hash-syntax
      if @@last.GetRegression()
        p_value = @@last.KSTest(fminer.all_activities.values, feat_hash.values).to_f # AM LAST: use internal function for test
        effect = (p_value > 0) ? "activating" : "deactivating"
      else
        p_value = @@last.ChisqTest(fminer.all_activities.values, feat_hash.values).to_f
        g=Array.new
        @value_map.each { |y,act| g[y-1]=Array.new }
        feat_hash.each  { |x,y|   g[y-1].push(x)   }
        max = OpenTox::Algorithm.effect(g, fminer.db_class_sizes)
        effect = max+1
      end
      feature_uri = File.join feature_dataset.uri,"feature","last", features.size.to_s
      unless features.include? smarts
        features << smarts
        metadata = {
          RDF.type => [OT.Feature, OT.Substructure],
          OT.hasSource => feature_dataset.uri,
          OT.smarts => smarts,
          OT.pValue => p_value.abs,
          OT.effect => effect,
          OT.parameters => [
            { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] },
            { DC.title => "prediction_feature", OT.paramValue => params[:prediction_feature] }
        ]
        }
        feature_dataset.add_feature feature_uri, metadata
      end
      if !hit_count
        ids.each { |id| feature_dataset.add(fminer.compounds[id], feature_uri, 1)}
      else
        ids.each_with_index { |id,i| feature_dataset.add(fminer.compounds[id], feature_uri, counts[smarts][i])}
      end
    end

    # AM: add feature values for non-present features
    # feature_dataset.complete_data_entries

    feature_dataset.save(@subjectid)
    feature_dataset.uri
  end
  response['Content-Type'] = 'text/uri-list'
  raise OpenTox::ServiceUnavailableError.newtask.uri+"\n" if task.status == "Cancelled"
  halt 202,task.uri.to_s+"\n"
end

# Matches features of a a feature dataset onto instances of another dataset. 
# The latter is referred to as 'training dataset', since p-values are computed,
# if user passes a prediction feature, or if the training dataset has only one feature.
# The result does not contain the prediction feature.
# @param [String] dataset_uri URI of the dataset 
# @param [String] feature_dataset_uri URI of the feature dataset (i.e. dependent variable)
# @param [optional] parameters Accepted parameters are
# - prediction_feature URI of prediction feature to calculate p-values for
# @return [text/uri-list] Task URI
post '/fminer/:method/match?' do
  raise OpenTox::BadRequestError.new "feature_dataset_uri not given" unless params[:feature_dataset_uri]
  raise OpenTox::BadRequestError.new "dataset_uri not given" unless params[:dataset_uri]

  training_dataset = OpenTox::Dataset.find "#{params[:dataset_uri]}",@subjectid
  unless params[:prediction_feature] # try to read prediction_feature from dataset
    prediction_feature = OpenTox::Feature.find(training_dataset.features.keys.first,@subjectid) if training_dataset.features.size == 1
  end
  prediction_feature = OpenTox::Feature.find(params[:prediction_feature],@subjectid) if params[:prediction_feature]

  task = OpenTox::Task.create("Matching features", url_for('/fminer/match',:full)) do |task|

    # get endpoint statistics
    if prediction_feature
      db_class_sizes = Array.new # for effect calculation
      all_activities = Hash.new # for effect calculation, indexed by id, starting from 1 (not 0)
      id = 1
      training_dataset.compounds.each do |compound|
        entry=training_dataset.data_entries[compound]
        entry.each do |feature,values|
          if feature == prediction_feature.uri
            values.each { |val|
              if val.nil? 
                LOGGER.warn "No #{feature} activity for #{compound.to_s}."
              else
                if prediction_feature.feature_type == "classification"
                  activity= training_dataset.value_map(prediction_feature.uri).invert[val].to_i # activities are mapped to 1..n
                  db_class_sizes[activity-1].nil? ? db_class_sizes[activity-1]=1 : db_class_sizes[activity-1]+=1 # AM effect
                elsif prediction_feature.feature_type == "regression"
                  activity= val.to_f 
                end
                begin
                  all_activities[id]=activity # DV: insert global information
                  id += 1
                rescue Exception => e
                  LOGGER.warn "Could not add " + smiles + "\t" + val.to_s + " to fminer"
                  LOGGER.warn e.backtrace
                end
              end
            }
          end
        end
      end
    end

    # Intialize result by adding compounds
    f_dataset = OpenTox::Dataset.find params[:feature_dataset_uri],@subjectid
    c_dataset = OpenTox::Dataset.find params[:dataset_uri],@subjectid
    res_dataset = OpenTox::Dataset.create CONFIG[:services]["dataset"],@subjectid
    c_dataset.compounds.each do |c|
      res_dataset.add_compound(c)
    end

    # Run matching, put data entries in result. Features are recreated.
    smi = [nil]; smi += c_dataset.compounds.collect { |c| OpenTox::Compound.new(c).to_smiles }
    smarts = f_dataset.features.collect { |f,m| m[OT.smarts] }
    params[:nr_hits] == "true" ? hit_count=true: hit_count=false
    matches, counts = LU.new.match_rb(smi, smarts, hit_count) if smarts.size>0

    f_dataset.features.each do |f,m|
      if (matches[m[OT.smarts]] && matches[m[OT.smarts]].size>0)

        feature_uri = File.join params[:feature_dataset_uri],"feature","bbrc","match", res_dataset.features.size.to_s
        #feature_uri = File.join res_dataset.uri,"feature","match", res_dataset.features.size.to_s
        metadata = {
          RDF.type => [OT.Feature, OT.Substructure],
          OT.hasSource => f_dataset.uri,
          OT.smarts => m[OT.smarts],
          OT.parameters => [
            { DC.title => "dataset_uri", OT.paramValue => params[:dataset_uri] }
          ]
        }

        if (prediction_feature) 
          feat_hash = Hash[*(all_activities.select { |k,v| matches[m[OT.smarts]].include?(k) }.flatten)]
          if prediction_feature.feature_type == "regression"
            p_value = @@last.KSTest(all_activities.values, feat_hash.values).to_f # AM LAST: use internal function for test
            effect = (p_value > 0) ? "activating" : "deactivating"
          else
            p_value = @@last.ChisqTest(all_activities.values, feat_hash.values).to_f
            g=Array.new # g is filled in *a*scending activity
            training_dataset.value_map(prediction_feature.uri).each { |y,act| g[y-1]=Array.new }
            feat_hash.each  { |x,y|   g[y-1].push(x)   }
            max = OpenTox::Algorithm.effect(g, db_class_sizes) # db_class_sizes is filled in *a*scending activity
            effect = max+1
          end
          metadata[OT.effect] = effect
          metadata[OT.pValue] = ((p_value.abs * 10000).round / 10000).to_f
          metadata[OT.parameters] << { DC.title => "prediction_feature", OT.paramValue => prediction_feature.uri }
        end
        
        res_dataset.add_feature feature_uri, metadata

        matches[m[OT.smarts]].each_with_index {|id,idx| 
          res_dataset.add(c_dataset.compounds[id-1],feature_uri,counts[m[OT.smarts]][idx])
        }
      end
    end
    res_dataset.save @subjectid
    res_dataset.uri
  end
  return_task(task)
end