def plot_cost(self, figure_id=None, new_figure=False, **kwargs):
r"""
Plots the cost evolution throughout the fitting.
"""
legend = self.algorithm
x_label = 'Number of iterations'
y_label = 'Normalized cost'
return GraphPlotter(figure_id, new_figure, range(0, self.n_iters+1),
[self.costs], legend=legend, x_label=x_label,
y_label=y_label).render(**kwargs)
def plot_error(self, figure_id=None, new_figure=False, **kwargs):
r"""
Plots the error evolution throughout the fitting.
"""
if self.gt_shape is not None:
legend = [self.algorithm]
x_label = 'Number of iterations'
y_label = self._error_text
return GraphPlotter(figure_id, new_figure,
range(0, self.n_iters+1), [self.errors],
legend=legend, x_label=x_label,
y_label=y_label).render(**kwargs)
else:
raise ValueError('Ground truth has not been set, error '
'cannot be plotted')
def _plot_dist(self, x_axis, y_axis, title, figure_id=None,
new_figure=False, y_limit=1, **kwargs):
legend = [self.algorithm +
'\nmean: {0:.4f}'.format(self.final_mean_error) +
'std: {0:.4f}, '.format(self.final_std_error) +
'median: {0:.4f}, '.format(self.final_median_error) +
'convergence: {0:.2f}, '.format(self.convergence),
'Initialization' +
'\nmean: {0:.4f}, '.format(self.initial_mean_error) +
'std: {0:.4f}, '.format(self.initial_std_error) +
'median: {0:.4f}, '.format(self.initial_median_error)]
x_label = self._error_text
y_label = 'Proportion of images'
axis_limits = [0, self._error_stop, 0, y_limit]
return GraphPlotter(figure_id, new_figure, x_axis, y_axis,
title=title, legend=legend,
x_label=x_label, y_label=y_label,
axis_limits=axis_limits).render(**kwargs)
def plot_cost(self, figure_id=None, new_figure=False, **kwargs):
r"""
Plots the cost evolution throughout the fitting.
"""
title = 'Cost evolution'
legend = self.algorithm
x_label = 'Number of iterations'
y_label = 'Normalized cost'
costs = [c for cost in self.costs for c in cost]
total_n_iters = self.n_iters + self.n_levels
axis_limits = [0, total_n_iters, 0, max(costs)]
return GraphPlotter(figure_id,
new_figure,
range(0, self.n_iters + self.n_levels),
costs,
title=title,
legend=legend,
x_label=x_label,
y_label=y_label,
axis_limits=axis_limits).render(**kwargs)
def plot_error(self, figure_id=None, new_figure=False, **kwargs):
r"""
Plots the error evolution throughout the fitting.
"""
if self.gt_shape is not None:
title = 'Error evolution'
legend = [self.algorithm]
x_label = 'Number of iterations'
y_label = self._error_text
errors = self.errors
x_limit = self.n_iters + self.n_levels
axis_limits = [0, x_limit, 0, np.max(errors)]
return GraphPlotter(figure_id,
new_figure,
range(0, x_limit), [errors],
title=title,
legend=legend,
x_label=x_label,
y_label=y_label,
axis_limits=axis_limits).render(**kwargs)
else:
raise ValueError('Ground truth shape has not been set, error '
'cannot be plotted')
def plot_fitting_curves(x_axis,
ceds,
title,
figure_id=None,
new_figure=False,
x_label='Point-to-Point Normalized RMS Error',
y_limit=1,
x_limit=0.05,
legend=None,
**kwargs):
r"""
Method that plots Cumulative Error Distributions in a single figure.
Parameters
----------
x_axis: ndarray
The horizontal axis values (errors).
ceds: list of ndarrays
The vertical axis values (percentages).
title: string
The plot title.
figure_id, Optional
A figure handle.
Default: None
new_figure: boolean, Optional
If True, a new figure window will be created.
Default: False
y_limit: float, Optional
The maximum value of the vertical axis.
Default: 1
x_limit: float, Optional
The maximum value of the vertical axis.
Default: 0.05
x_label: string
The label of the horizontal axis.
Default: 'Point-to-Point Normalized RMS Error'
legend: list of strings or None
The legend of the plot. If None, the legend will include an incremental
number per curve.
Default: None
Returns
-------
final_error_dist: list
Cumulative distribution values of the final errors.
initial_error_dist: list
Cumulative distribution values of the initial errors.
"""
if legend is None:
legend = [str(i + 1) for i in range(len(ceds))]
y_label = 'Proportion of images'
axis_limits = [0, x_limit, 0, y_limit]
return GraphPlotter(figure_id,
new_figure,
x_axis,
ceds,
title=title,
legend=legend,
x_label=x_label,
y_label=y_label,
axis_limits=axis_limits).render(**kwargs)