def plot_marginals(self, selector=None, axes=None):
"""Visualizes the marginal posterior distributions for each parameter.
Parameters
----------
selector : iterable of ints or strings, optional
Indices or keys to use from marginals. Default to all.
axes: one or an iterable of plt.Axes, optional
Returns
-------
axes: np.ndarray of plt.Axes
"""
# TODO: allow kwargs
marginals = self.marginals
unique_param_vals = self._get_unique_parameter_values()
ncols = len(marginals.keys()) if len(marginals.keys()) > 5 else 5
marginals = viz._limit_params(marginals, selector)
shape = (max(1, len(marginals) // ncols), min(len(marginals), ncols))
axes, _ = viz._create_axes(axes, shape)
axes = axes.ravel()
for idx, key in enumerate(marginals.keys()):
axes[idx].plot(unique_param_vals[key], marginals[key])
axes[idx].fill_between(unique_param_vals[key],
marginals[key],
alpha=0.1)
axes[idx].set_xlabel(key)
return axes
def plot_pairs(self, selector=None, axes=None):
"""Visualizes pairwise relationships as a matrix with marginals on the diagonal.
Parameters
----------
selector: iterable of ints or strings, optional
Indices or keys to use from marginals and posterior. Default to all.
axes: one or an iterable of plt.Axes, optional
Returns
-------
axes: np.ndarray of plt.Axes
"""
# TODO: allow kwargs
posterior_shape = self._get_number_of_unique_parameter_values()
posterior = self._posterior.reshape(posterior_shape) / np.sum(
self._posterior)
marginals = self.marginals
unique_param_vals = self._get_unique_parameter_values()
marginals = viz._limit_params(marginals, selector)
shape = (len(marginals), len(marginals))
axes, _ = viz._create_axes(axes, shape)
for idx_row, key_row in enumerate(marginals):
for idx_col, key_col in enumerate(marginals):
if idx_row == idx_col:
# plot 1d marginals
axes[idx_row, idx_col].plot(unique_param_vals[key_row],
marginals[key_row])
axes[idx_row,
idx_col].fill_between(unique_param_vals[key_row],
marginals[key_row],
alpha=0.1)
else:
# plot 2d marginals
xx, yy = np.meshgrid(unique_param_vals[key_col],
unique_param_vals[key_row],
indexing='ij')
axes[idx_row, idx_col].contourf(*[xx, yy],
self._get_2d_marginal(
idx_row, idx_col,
posterior),
cmap='Blues')
axes[idx_row, 0].set_ylabel(key_row)
axes[-1, idx_row].set_xlabel(key_row)
return axes
def plot_discrepancy(self, axes=None, **kwargs):
"""Plot acquired parameters vs. resulting discrepancy.
TODO: refactor
"""
n_plots = self.target_model.input_dim
ncols = kwargs.pop('ncols', 5)
kwargs['sharey'] = kwargs.get('sharey', True)
shape = (max(1, n_plots // ncols), min(n_plots, ncols))
axes, kwargs = vis._create_axes(axes, shape, **kwargs)
axes = axes.ravel()
for ii in range(n_plots):
axes[ii].scatter(self.target_model._gp.X[:, ii], self.target_model._gp.Y[:, 0])
axes[ii].set_xlabel(self.parameter_names[ii])
axes[0].set_ylabel('Discrepancy')
return axes