m = lambda x: model.predict(x)
sd2 = lambda x: model.predict(x, eval_MSE=True)[1]
m_dx = lambda x: model.gradient(x)[0]
sd2_dx = lambda x: model.gradient(x)[1]
if 1 < 2:
fig0, (ax0, ax1, ax2) = plt.subplots(1, 3, sharey=False, sharex=False,
figsize=(fig_width, fig_height),
subplot_kw={'aspect': 'equal'}, dpi=100)
gs1 = gridspec.GridSpec(1, 3)
gs1.update(wspace=0.025, hspace=0.05) # set the spacing between axes.
plot_contour_gradient(ax0, fitness, None, x_lb, x_ub, title='Noisy function',
n_level=20, n_per_axis=200)
plot_contour_gradient(ax1, m, m_dx, x_lb, x_ub, title='GPR estimation',
n_level=20, n_per_axis=200)
plot_contour_gradient(ax2, sd2, sd2_dx, x_lb, x_ub, title='GPR variance',
n_level=20, n_per_axis=200)
plt.tight_layout()
fig1, ax3 = plt.subplots(1, 1, figsize=(fig_width, fig_height),
subplot_kw={'aspect': 'equal'}, dpi=100)
plot_contour_gradient(ax3, infill, infill_dx, x_lb, x_ub, title='Infill-Criterion',
is_log=True, n_level=50, n_per_axis=250)
plt.tight_layout()
f = EI(model)
def __(x):
_, dx = f(x, dx=True)
return dx
return __
f = EI(model)
dx = grad(model)
fig0, ax0 = plt.subplots(1,
1,
sharey=True,
sharex=False,
figsize=(fig_width, fig_height),
subplot_kw={'aspect': 'auto'},
dpi=100)
plot_contour_gradient(ax0,
f,
dx,
x_lb,
x_ub,
title='Function',
n_level=15,
n_per_axis=120)
plt.tight_layout()
plt.show()