649a569555396206384c92a7e2a235f513bfc3be,geomstats/learning/frechet_mean.py,,_adaptive_gradient_descent,#Any#Any#Any#Any#Any#Any#Any#,177

Before Change

    tau_mul_up = 1.6511111
    tau_min = 1e-6
    tau_mul_down = 0.1
    if point_type == "matrix":
        raise NotImplementedError(
            "The Frechet mean with adaptive gradient descent is only"
            " implemented for lists of vectors, and not matrices.")

    n_points = 1
    if gs.ndim(points) == 2:
        n_points = gs.shape(points)[0]

    if n_points == 1:
        return points

    if weights is None:
        weights = gs.ones((n_points,))

    //weights = gs.array(weights)
    //weights = gs.to_ndarray(weights, to_ndim=2, axis=1)

    sum_weights = gs.sum(weights)

    if init_point is None:
        current_mean = points[0]
    else:
        current_mean = init_point

    tau = 1.0
    iteration = 0

    logs = metric.log(point=points, base_point=current_mean)
    current_tangent_mean = gs.einsum("...,...j->j", weights, logs)
    current_tangent_mean /= sum_weights
    sq_norm_current_tangent_mean = metric.squared_norm(
        current_tangent_mean, base_point=current_mean)

    while (sq_norm_current_tangent_mean > epsilon ** 2
           and iteration < max_iter):
        iteration = iteration + 1
        shooting_vector = gs.to_ndarray(
            tau * current_tangent_mean,
            to_ndim=2)
        next_mean = metric.exp(
            tangent_vec=shooting_vector,
            base_point=current_mean)
        logs = metric.log(point=points, base_point=next_mean)

After Change

    current_mean: array-like, shape=[n_samples, dim]
        Weighted Frechet mean of the points.
    
    if point_type == "matrix":
        raise NotImplementedError(
            "The Frechet mean with adaptive gradient descent is only"
            " implemented for lists of vectors, and not matrices.")

    tau_max = 1e6
    tau_mul_up = 1.6511111
    tau_min = 1e-6
    tau_mul_down = 0.1

    n_points = geomstats.vectorization.get_n_points(
        points, point_type)

    points = gs.to_ndarray(points, to_ndim=2)

    if n_points == 1:
        return points[0]

    if weights is None:
        weights = gs.ones((n_points,))

    //weights = gs.array(weights)
    //weights = gs.to_ndarray(weights, to_ndim=2, axis=1)

    sum_weights = gs.sum(weights)

    current_mean = points[0] if init_point is None else init_point

    tau = 1.0
    iteration = 0

    logs = metric.log(point=points, base_point=current_mean)
    current_tangent_mean = gs.einsum("...,...j->j", weights, logs)
    current_tangent_mean /= sum_weights
    sq_norm_current_tangent_mean = metric.squared_norm(
        current_tangent_mean, base_point=current_mean)

    print("init")
    print(current_tangent_mean.shape)
    while (sq_norm_current_tangent_mean > epsilon ** 2
           and iteration < max_iter):
        iteration += 1

        shooting_vector = tau * current_tangent_mean
        next_mean = metric.exp(
            tangent_vec=shooting_vector, base_point=current_mean)

        logs = metric.log(point=points, base_point=next_mean)
        next_tangent_mean = gs.einsum("...,...j->j", weights, logs)
        next_tangent_mean /= sum_weights
        sq_norm_next_tangent_mean = metric.squared_norm(
            next_tangent_mean, base_point=next_mean)

        if sq_norm_next_tangent_mean < sq_norm_current_tangent_mean:
            current_mean = next_mean
            current_tangent_mean = next_tangent_mean
            sq_norm_current_tangent_mean = sq_norm_next_tangent_mean
            tau = min(tau_max, tau_mul_up * tau)
        else:
            tau = max(tau_min, tau_mul_down * tau)
        print(current_tangent_mean.shape)

    if iteration == max_iter:
        logging.warning(
            "Maximum number of iterations {} reached. "

Instances