Function quantize_to_u8_bits

src/models/common.rs:238–275

(fins: &[f32], resolution: u8)

}

pub fn quantize_to_u8_bits(fins: &[f32], resolution: u8) -> Vec<Vec<u8>> {
    let bits_per_value = resolution as usize;
    let parts = 2_usize.pow(bits_per_value as u32);
    let step = 2.0 / parts as f32;
    let u8s_per_value = (fins.len() + 7) / 8;
    let mut quantized: Vec<Vec<u8>> = vec![Vec::with_capacity(u8s_per_value); bits_per_value];

    let mut current_u8s: Vec<u8> = vec![0; bits_per_value];
    let mut bit_index: usize = 0;

    for &f in fins {
        let flags = to_float_flag(f, bits_per_value, step);

        for bit_position in 0..bits_per_value {
            if flags[bit_position] {
                current_u8s[bit_position] |= 1 << bit_index;
            }
        }
        bit_index += 1;

        if bit_index == 8 {
            for bit_position in 0..bits_per_value {
                quantized[bit_position].push(current_u8s[bit_position]);
                current_u8s[bit_position] = 0;
            }
            bit_index = 0;
        }
    }

    // Push remaining bits if not a multiple of 8
    if bit_index > 0 {
        for bit_position in 0..bits_per_value {
            quantized[bit_position].push(current_u8s[bit_position]);
        }
    }

    quantized
}

#[derive(Debug, Clone)]
pub enum WaCustomError {