expertiza/expertiza

# new class created to handle code related to cycles between reviews
# This code was previously in AssignmentParticipant and is not being used anywhere yet
class CollusionCycle
  # Cycle data structure
  # Each edge of the cycle stores a participant and the score given by to the participant by the reviewer.
  # Consider a 3 node cycle: A --> B --> C --> A (A reviewed B; B reviewed C and C reviewed A)
  # For the above cycle, the data structure would be: [[A, SCA], [B, SAB], [C, SCB]], where SCA is the score given by C to A.

  def two_node_cycles(assignment_participant)
    collusion_cycles = []
    assignment_participant.reviewers.each do |ap|
      next unless ap.reviewers.include?(assignment_participant)
      if assignment_participant.reviews_by_reviewer(ap).nil?
        next
      else
        s01 = assignment_participant.reviews_by_reviewer(ap).aggregate_questionnaire_score
      end

      if ap.reviews_by_reviewer(assignment_participant).nil?
        next
      else
        s10 = ap.reviews_by_reviewer(assignment_participant).aggregate_questionnaire_score
      end

      collusion_cycles.push([[assignment_participant, s01], [ap, s10]])
    end
    collusion_cycles
  end

  def three_node_cycles(assignment_participant)
    collusion_cycles = []
    assignment_participant.reviewers.each do |ap1|
      ap1.reviewers.each do |ap2|
        next unless ap2.reviewers.include?(assignment_participant)
        if assignment_participant.reviews_by_reviewer(ap1).nil?
          next
        else
          s01 = assignment_participant.reviews_by_reviewer(ap1).aggregate_questionnaire_score
        end

        if ap1.reviews_by_reviewer(ap2).nil?
          next
        else
          s12 = ap1.reviews_by_reviewer(ap2).aggregate_questionnaire_score
        end

        if ap2.reviews_by_reviewer(assignment_participant).nil?
          next
        else
          s20 = ap2.reviews_by_reviewer(assignment_participant).aggregate_questionnaire_score
        end

        collusion_cycles.push([[assignment_participant, s01], [ap1, s12], [ap2, s20]])
      end
    end

    collusion_cycles
  end

  def four_node_cycles(assignment_participant)
    collusion_cycles = []
    assignment_participant.reviewers.each do |ap1|
      ap1.reviewers.each do |ap2|
        ap2.reviewers.each do |ap3|
          next unless ap3.reviewers.include?(assignment_participant)

          if assignment_participant.reviews_by_reviewer(ap1).nil?
            next
          else
            s01 = assignment_participant.reviews_by_reviewer(ap1).aggregate_questionnaire_score
          end

          if ap1.reviews_by_reviewer(ap2).nil?
            next
          else
            s12 = ap1.reviews_by_reviewer(ap2).aggregate_questionnaire_score
          end

          if ap2.reviews_by_reviewer(ap3).nil?
            next
          else
            s23 = ap2.reviews_by_reviewer(ap3).aggregate_questionnaire_score
          end

          if ap3.reviews_by_reviewer(assignment_participant).nil?
            next
          else
            s30 = ap3.reviews_by_reviewer(assignment_participant).aggregate_questionnaire_score
          end

          collusion_cycles.push([[assignment_participant, s01], [ap1, s12], [ap2, s23], [ap3, s30]])
        end
      end
    end
    collusion_cycles
  end

  # Per cycle
  def cycle_similarity_score(cycle)
    similarity_score = 0.0
    count = 0.0
    (0...cycle.size - 1).each do |pivot|
      pivot_score = cycle[pivot][1]
      (pivot + 1...cycle.size).each do |other|
        similarity_score += (pivot_score - cycle[other][1]).abs
        count += 1.0
      end
    end
    similarity_score /= count unless count == 0.0
    similarity_score
  end

  # Per cycle
  def cycle_deviation_score(cycle)
    deviation_score = 0.0
    count = 0.0
    (0...cycle.size).each do |member|
      participant = AssignmentParticipant.find(cycle[member][0].id)
      total_score = participant.review_score
      deviation_score += (total_score - cycle[member][1]).abs
      count += 1.0
    end
    deviation_score /= count unless count == 0.0
    deviation_score
  end
end