Skip to content

Commit

Permalink
hotfix for using matplotlib
Browse files Browse the repository at this point in the history
  • Loading branch information
@ChrisLisbon
ChrisLisbon committed Dec 15, 2023
1 parent 8d59e90 commit b89c253
Showing 1 changed file with 39 additions and 35 deletions.
74 changes: 39 additions & 35 deletions epde/optimizers/moeadd/vis.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,88 +5,92 @@
import matplotlib.pyplot as plt
import matplotlib as mpl

mpl.rcParams.update(mpl.rcParamsDefault)
plt.rcParams['text.usetex'] = True

SMALL_SIZE = 12
mpl.rc('font', size=SMALL_SIZE)
mpl.rc('axes', titlesize=SMALL_SIZE)

class ParetoVisualizer(object):
def __init__(self, eq_pareto_levels):
'''
"""
Проще всего получить pareto_levels из атрибута optimizer.pareto_levels
'''
"""
self.pareto_frontier = eq_pareto_levels

def plot_pareto(self, dimensions: tuple = (0, 1), annotate_best=True, plot_level = 1,
filename = None, save_format = 'eps'):
def plot_pareto(self, dimensions: tuple = (0, 1), annotate_best=True, plot_level=1,
filename=None, save_format='eps'):

#TODO replace latex with Mathtext
mpl.rcParams.update(mpl.rcParamsDefault)
plt.rcParams['text.usetex'] = True

SMALL_SIZE = 12
mpl.rc('font', size=SMALL_SIZE)
mpl.rc('axes', titlesize=SMALL_SIZE)

assert len(
dimensions) == 2, 'The pareto levels are projected on the 2D plane, therefore only 2 coordinates are processible'
coords = [[(solution.obj_fun[dimensions[0]], solution.obj_fun[dimensions[1]])
coords = [[(solution.obj_fun[dimensions[0]], solution.obj_fun[dimensions[1]])
for solution in self.pareto_frontier.levels[front_idx]]
for front_idx in np.arange(plot_level)]#len(self.pareto_frontier.levels))]
for front_idx in np.arange(plot_level)] # len(self.pareto_frontier.levels))]
if annotate_best:
try:
annotations = [[solution.latex_form for solution in self.pareto_frontier.levels[front_idx]]
for front_idx in np.arange(len(self.pareto_frontier.levels))]
except AttributeError:
annotations = [[str(solution.obj_fun) for solution in self.pareto_frontier.levels[front_idx]]
for front_idx in np.arange(len(self.pareto_frontier.levels))]
for front_idx in np.arange(len(self.pareto_frontier.levels))]

coords_arrays = []
for coord_set in coords:
coords_arrays.append(np.array(coord_set))

colors = ['r', 'k', 'b', 'y', 'g'] + \
['m' for idx in np.arange(len(coords_arrays) - 5)]
['m' for idx in np.arange(len(coords_arrays) - 5)]

if len(coords_arrays) > 1:
x_min = min(*[np.min(coord_arr[:, 0]) for coord_arr in coords_arrays])
x_max = max(*[np.max(coord_arr[:, 0]) for coord_arr in coords_arrays])
y_min = min(*[np.min(coord_arr[:, 1]) for coord_arr in coords_arrays])
y_max = max(*[np.max(coord_arr[:, 1]) for coord_arr in coords_arrays])
else:
x_min = np.min(coords_arrays[0][:, 0]); x_max = np.max(coords_arrays[0][:, 0])
y_min = np.min(coords_arrays[0][:, 1]); y_max = np.max(coords_arrays[0][:, 1])

x_min = np.min(coords_arrays[0][:, 0]);
x_max = np.max(coords_arrays[0][:, 0])
y_min = np.min(coords_arrays[0][:, 1]);
y_max = np.max(coords_arrays[0][:, 1])

x_interval = max(x_max - x_min, 5)
y_interval = max(y_max - y_min, 2)

plt.grid()
plt.xlim(x_min - 0.1 * x_interval, x_max + 0.8 * x_interval) # ax.set_
plt.ylim( y_min - 0.1 * y_interval, y_max + 0.3 * y_interval) # ax.set_
plt.xlim(x_min - 0.1 * x_interval, x_max + 0.8 * x_interval) # ax.set_
plt.ylim(y_min - 0.1 * y_interval, y_max + 0.3 * y_interval) # ax.set_

plt.xlabel('Objective 1')
plt.ylabel('Objective 2')

for front_idx in np.arange(min(len(coords_arrays), plot_level)):

plt.scatter(coords_arrays[front_idx][:, 0],
coords_arrays[front_idx][:, 1], color=colors[front_idx])
coords_arrays[front_idx][:, 1], color=colors[front_idx])
locs_used = []
for front_elem_idx in np.arange(coords_arrays[front_idx].shape[0]):
if any([all(np.isclose(np.array((coords_arrays[front_idx][front_elem_idx, 0],
if any([all(np.isclose(np.array((coords_arrays[front_idx][front_elem_idx, 0],
coords_arrays[front_idx][front_elem_idx, 1])), entry))
for entry in locs_used]):
continue
else:
locs_used.append(np.array((coords_arrays[front_idx][front_elem_idx, 0],
locs_used.append(np.array((coords_arrays[front_idx][front_elem_idx, 0],
coords_arrays[front_idx][front_elem_idx, 1])))
if front_idx == 0 or not annotate_best:
if annotate_best:
annotation = annotations[front_idx][front_elem_idx]
if annotation[0] != r'$':
annotation = r'$' + annotation + r'$'
print(annotation)
plt.annotate(annotations[front_idx][front_elem_idx],
(coords_arrays[front_idx][front_elem_idx, 0] + 0.4,
coords_arrays[front_idx][front_elem_idx, 1] + 0.2),
bbox = dict(boxstyle="Square,pad=0.3",
fc="white", lw=0.5))#,
# fontsize = 'xx-large')

plt.annotate(annotations[front_idx][front_elem_idx],
(coords_arrays[front_idx][front_elem_idx, 0] + 0.4,
coords_arrays[front_idx][front_elem_idx, 1] + 0.2),
bbox=dict(boxstyle="Square,pad=0.3",
fc="white", lw=0.5)) # ,
# fontsize = 'xx-large')

if filename is not None:
plt.savefig(filename + '.' + save_format, dpi = 300, quality = 94, format=save_format)
plt.savefig(filename + '.' + save_format, dpi=300, quality=94, format=save_format)
plt.show()

0 comments on commit b89c253

Please sign in to comment.