giuse/machine_learning_workbench

# frozen_string_literal: true

module MachineLearningWorkbench::NeuralNetwork
  # Recurrent Neural Network
  class Recurrent < Base

    # Calculate the size of each row in a layer's weight matrix.
    # Each row holds the inputs for the next level: previous level's
    # activations (or inputs), this level's last activations
    # (recursion) and bias.
    # @return [Array<Integer>] per-layer row sizes
    def layer_row_sizes
      @layer_row_sizes ||= struct.each_cons(2).collect do |prev, rec|
        prev + rec + 1
      end
    end

      # # NOTE: current layer index corresponds to index of next state!
      # previous = nlay     # index of previous layer (inputs)
      # current = nlay + 1  # index of current layer (outputs)
      # # Copy the level's last-time activation to the input (previous state)
      # # TODO: ranges in `NArray#[]` should be reliable, get rid of loop
      # nneurs(current).times do |i| # for each activations to copy
      #   # Copy output from last-time activation to recurrency in previous state
      #   @state[previous][0, nneurs(previous) + i] = state[current][0, i]
      # end
      # act_fn.call state[previous].dot layers[nlay]

    # Activates a layer of the network.
    # Bit more complex since it has to copy the layer's activation on
    # last input to its own inputs, for recursion.
    # @param i [Integer] the layer to activate, zero-indexed
    def activate_layer nlay
      # Mark begin and end of recursion outputs in current state
      begin_recur = nneurs(nlay)
      end_recur = nneurs(nlay) + nneurs(nlay+1)
      # Copy the level's last-time activation to the current input recurrency
      state[nlay][begin_recur...end_recur] = state[nlay+1][0...nneurs(nlay+1)]
      # Activate current layer
      act_fn.call state[nlay].dot layers[nlay]
    end

  end
end