plot_3d_gp¶

safeopt.plot_3d_gp(gp, inputs, predictions=None, figure=None, axis=None, fixed_inputs=None, beta=3, **kwargs)¶

Plot a 3D gp with uncertainty.

Parameters:

gp: Instance of GPy.models.GPRegression
inputs: 2darray: The input parameters at which the GP is to be evaluated
predictions: ndarray: Can be used to manually pass the GP predictions, set to None to use the gp directly. Is of the form [mean, variance]
figure: matplotlib figure: The figure on which to draw (ignored if axis is provided
axis: matplotlib axis: The axis on which to draw
fixed_inputs: list: A list containing the the fixed inputs and their corresponding values, e.g., [(0, 3.2), (4, -2.43)]. Set the value to None if it’s not fixed, but should not be a plotted axis either
beta: float: The confidence interval used

Returns: