Function gpfa_m_step 

fn gpfa_m_step(
    y: &[f64],
    x_post: &[f64],
    xx_post: &[f64],
    n_neurons: usize,
    n_bins: usize,
    n_latents: usize,
) -> (Vec<f64>, Vec<f64>, Vec<f64>)

M-step: update C, d, R.