Function generate_data

doc/sphinx/notebooks/tailor/external-mean-function.pct.py:48–63

For each function, sample the value at `N` equally spaced points in the [−5, 5] interval (Fortuin and Rätsch, 2019). Returns: Tuple of np.arrays of size (N, 1) and (N, num_functions).

(num_functions=10, N=500)

# %%
def generate_data(num_functions=10, N=500):
    """
    For each function, sample the value at `N` equally spaced
    points in the [−5, 5] interval (Fortuin and Rätsch, 2019).

    Returns:
        Tuple of np.arrays of size (N, 1) and (N, num_functions).
    """
    jitter = 1e-6
    Xs = np.linspace(-5.0, 5.0, N)[:, None]
    kernel = RBF(lengthscales=1.0)
    cov = kernel(Xs)
    L = np.linalg.cholesky(cov + np.eye(N) * jitter)
    epsilon = np.random.randn(N, num_functions)
    F = np.sin(Xs) + np.matmul(L, epsilon)
    return Xs, F


# %% [markdown]