Class: JXL::Modular::Weighted
- Inherits:
-
Object
- Object
- JXL::Modular::Weighted
- Defined in:
- lib/jxl/modular/weighted.rb
Constant Summary collapse
- DIV_LOOKUP =
Array.new(64) { |i| (1 << 24) / (i + 1) }.freeze
Instance Attribute Summary collapse
-
#property ⇒ Object
readonly
Returns the value of attribute property.
Instance Method Summary collapse
-
#initialize(header, width) ⇒ Weighted
constructor
A new instance of Weighted.
-
#predict(plane, x, y) ⇒ Object
rubocop:disable Metrics/AbcSize.
- #update(value, x) ⇒ Object
Constructor Details
#initialize(header, width) ⇒ Weighted
Returns a new instance of Weighted.
10 11 12 13 14 15 16 |
# File 'lib/jxl/modular/weighted.rb', line 10 def initialize(header, width) @header = header @width = width @stride = width + 2 @pred_errors = Array.new(4) { Array.new(@stride * 2, 0) } @errors = Array.new(@stride * 2, 0) end |
Instance Attribute Details
#property ⇒ Object (readonly)
Returns the value of attribute property.
8 9 10 |
# File 'lib/jxl/modular/weighted.rb', line 8 def property @property end |
Instance Method Details
#predict(plane, x, y) ⇒ Object
rubocop:disable Metrics/AbcSize
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 |
# File 'lib/jxl/modular/weighted.rb', line 18 def predict(plane, x, y) # rubocop:disable Metrics/AbcSize n = Predictor.neighbors(plane, x, y) positions(x, y) weights = @pred_errors.each_with_index.map do |errors, i| sum = Num.u32(errors[@pos_n] + errors[@pos_ne] + errors[@pos_nw]) error_weight(sum, @header.weights[i]) end north, west, northeast, northwest, northnorth = %i[top left topright topleft toptop].map { n.fetch(_1) << 3 } error_west = x.zero? ? 0 : @errors[@current_row + x - 1] error_north = @errors[@pos_n] error_northwest = @errors[@pos_nw] error_northeast = @errors[@pos_ne] @property = [error_west, error_north, error_northwest, error_northeast].max_by(&:abs) sum = error_north + error_west p = @header.p @predictions = [ west + northeast - north, north - (((sum + error_northeast) * p[0]) >> 5), west - (((sum + error_northwest) * p[1]) >> 5), north - (((error_northwest * p[2]) + (error_north * p[3]) + (error_northeast * p[4]) + ((northnorth - north) * p[5]) + ((northwest - west) * p[6])) >> 5) ] @prediction = weighted_average(weights) same_sign = ((error_north ^ error_west) | (error_north ^ error_northwest)).positive? return (@prediction + 3) >> 3 if same_sign @prediction = @prediction.clamp([west, northeast, north].min, [west, northeast, north].max) (@prediction + 3) >> 3 end |
#update(value, x) ⇒ Object
51 52 53 54 55 56 57 58 59 |
# File 'lib/jxl/modular/weighted.rb', line 51 def update(value, x) scaled = value << 3 @errors[@current_row + x] = Num.i32(@prediction - scaled) @pred_errors.each_with_index do |errors, i| error = Num.u32(((@predictions[i] - scaled).abs + 3) >> 3) errors[@current_row + x] = error errors[@previous_row + x + 1] = Num.u32(errors[@previous_row + x + 1] + error) end end |