require File.expand_path( File.dirname(__FILE__) + '/../spec_helper' )
require File.expand_path( File.dirname(__FILE__) + '/../validator_helper' )


require 'validator/aa'
require 'spec_id'
require 'spec_id/digestor'

klass = Validator::AA

class MyAA < Validator::AA ; def initialize ; end ; end

describe klass, "using frequency estimates" do
  before(:each) do
    @small_fasta = Tfiles + '/small.fasta'
    @sequest_params = Tfiles + '/bioworks32.params'
    #         C/D      C/D    J     (7)
    @seqs = %w(ABCDEFGC CCDCCC JJJJJ XYZ WXXXYZ TXXXXXYZ ZZXIIPTYZ ZZXTYZZ ZZZZ YYYYYTL)
    @peps =  @seqs.map {|n| v = SpecID::GenericPep.new; v.aaseq = n ; v }
    val = klass.new('C')
    val.frequency = 0.11
    @validator = val
  end
  #C: 0.0157714433456144
  #D: 0.0526145691939758

  it_should_behave_like 'a validator'

  it 'calculates precision from actual and expected correctly' do
    MyAA.new.pephit_precision_from_actual_and_expected(5, 10, 100).should == 0.5
    (0..10).each do |actual|
      precision = MyAA.new.pephit_precision_from_actual_and_expected(actual, 10, 100)
      answer = 1.0 - (actual.to_f/10.0)
      precision.should be_close(answer, 0.00000001)
      #0 -> 1
      #1 -> 0.9
      #2 -> 0.8
    end
  end

  it 'calculates precision (with background) reasonably and consistently' do
    (0..10).each do |bkg|
      precision = MyAA.new.pephit_precision_from_actual_and_expected(5, 10, 100, bkg.to_f/100)
      answer = 0.5 + bkg.to_f/10
      answer = 1.0 if answer > 1.0 
      precision.should == answer
    end
  end

  it 'calculates at_least_one correctly' do
    aa = 'C'
    freq = 0.0157714433456144  ## from @small_fasta
    (actual, expected) = MyAA.new.at_least_one(aa, freq, @seqs)
    actual.should == 2
    expected.should be_close(0.949318337979434, 0.0001)  # freeze for consistency
  end

  it 'gives consistent precision of peptides given fastafile and aa (even negative)' do
    aa = 'C'
    val = klass.new(aa).set_frequency(Fasta.new(@small_fasta))
    # I checked this answer out by hand and it is correct
    val.pephit_precision(@peps).should be_close(-1.10677, 0.001)
  end
end

describe klass, "using empirical digestion data" do

  before(:each) do
    @small_fasta = Tfiles + '/small.fasta'
    @sequest_params = Tfiles + '/bioworks32.params'
    #         C/D      C/D    J     (7)
    @seqs = %w(ABCDEFGC CCDCCC JJJJJ XYZ WXXXYZ TXXXXXYZ ZZXIIPTYZ ZZXTYZZ ZZZZ YYYYYTL)
    @peps =  @seqs.map {|n| v = SpecID::GenericPep.new; v.aaseq = n ; v }

    val = klass.new('C')
    val.false_to_total_ratio = 0.22  # arbitrary
    @validator = val
  end

  it_should_behave_like 'a validator'

  it 'gives correct false to total ratio' do
    aa = 'C'
    val = klass.new(aa)
    peptides = Digestor.digest( Fasta.new(@small_fasta), Sequest::Params.new(@sequest_params))
    val.set_false_to_total_ratio( peptides )
    # frozen (but I checked the peptides by hand to make sure they were
    # correct)
    val.false_to_total_ratio.should be_close(0.177629264861062, 0.0000000000001)
  end

  it 'can validate with theoretical digestion or frequency estimate' do
    aa = 'C'
    val = klass.new(aa)
    val.false_to_total_ratio = 0.177629264861062
    answ = val.pephit_precision(@peps)
    # frozen
    answ.should be_close(-0.125940594059407, 0.0000000001)

    val.frequency = 0.0157714433456144  ## from @small_fasta
    val.false_to_total_ratio = nil
    answ = val.pephit_precision(@peps)
    answ.should be_close(-1.10677485094924, 0.0000000001)
  end
end