Method create_params

src/nn.rs:35–40

Creates random parameters which connect one layer with the other layer. The parameter `rows` denotes the number of layers in the right layer. The parameter `cols` denotes the number of layers in the left layer. If `from_input_layer` is `true` the left layer is an input layer. Otherwise, the left layer is a hidden layer.

(&self, rows: usize, cols: usize, from_input_layer: bool)

    /// `from_input_layer` is `true` the left layer is an input layer. Otherwise,
    /// the left layer is a hidden layer.
    fn create_params(&self, rows: usize, cols: usize, from_input_layer: bool) -> Matrix<f64> {

        let n = if from_input_layer { cols } else { cols + 1 };

        Matrix::from_vec(random::<f64>(rows * n), rows, n).unwrap()
    }

    /// Adds an layer to the neural network.
    pub fn add_layer(&self, n: usize) -> NeuralNetwork {