"""Plot summaries for the classifiers."""
from typing import List
import numpy as np
import matplotlib.pyplot as plt
from ..metrics import get_probs_prep_meas
from ..classifiers import Classifier, TwoStateClassifier, ThreeStateClassifier
from .shots1d import plot_two_pdfs_projected, plot_several_pdfs_along_line
from .shots2d import plot_shots_2d, plot_boundaries_2d, plot_contour_pdf_2d
from .metrics import plot_probs_prep_meas
[docs]
def summary(classifier: Classifier, *shots: List[np.ndarray]) -> plt.Figure:
"""Figure to show a general overview of the performance of the classifier."""
if classifier._num_states == 2:
fig = two_state_classifier(classifier, *shots)
elif classifier._num_states == 3:
fig = three_state_classifier(classifier, *shots)
else:
raise ValueError(
"Only implemented for two- and three-state classifiers, "
f"but {classifier._num_states}-state classifier was given. "
)
return fig
def two_state_classifier(
classifier: TwoStateClassifier,
shots_0: np.ndarray,
shots_1: np.ndarray,
) -> plt.Figure:
fig, axes = plt.subplots(nrows=1, ncols=3, figsize=(16, 5))
axes[0] = plot_shots_2d(axes[0], shots_0, shots_1)
probs = get_probs_prep_meas(classifier, shots_0, shots_1)
axes[1] = plot_probs_prep_meas(axes[1], probs)
axes[2] = plot_two_pdfs_projected(
axes[2],
classifier,
shots_0,
shots_1,
)
axes[2].set_title("projection: 0-1")
fig.tight_layout()
return fig
def three_state_classifier(
classifier: ThreeStateClassifier,
shots_0: np.ndarray,
shots_1: np.ndarray,
shots_2: np.ndarray,
) -> plt.Figure:
fig, axes = plt.subplots(nrows=2, ncols=3, figsize=(12, 7))
# (1) 2D plot with the experimental shots
plot_shots_2d(axes[0, 0], shots_0, shots_1, shots_2)
# (2) p(m|s) matrix
probs_prep_meas = get_probs_prep_meas(classifier, shots_0, shots_1, shots_2)
plot_probs_prep_meas(axes[0, 1], probs_prep_meas)
# (3) 2D plot with decision boundaries and 1/e countour
plot_shots_2d(axes[0, 2], shots_0, shots_1, shots_2)
plot_boundaries_2d(axes[0, 2], classifier)
plot_contour_pdf_2d(axes[0, 2], classifier.pdf_0)
plot_contour_pdf_2d(axes[0, 2], classifier.pdf_1)
plot_contour_pdf_2d(axes[0, 2], classifier.pdf_2)
axes[0, 2].legend().remove()
# (4-6) 1D plot along a line
params = classifier.params
mu_0 = [params[0]["mu_0_x"], params[0]["mu_0_y"]]
mu_1 = [params[1]["mu_1_x"], params[1]["mu_1_y"]]
mu_2 = [params[2]["mu_2_x"], params[2]["mu_2_y"]]
plot_several_pdfs_along_line(
axes[1, 0],
[mu_0, mu_1],
classifier,
shots_0,
shots_1,
shots_2,
)
axes[1, 0].set_title("projection: 0-1")
axes[1, 0].legend().remove()
plot_several_pdfs_along_line(
axes[1, 1],
[mu_1, mu_2],
classifier,
shots_0,
shots_1,
shots_2,
)
axes[1, 1].set_title("projection: 1-2")
axes[1, 1].legend().remove()
plot_several_pdfs_along_line(
axes[1, 2],
[mu_0, mu_2],
classifier,
shots_0,
shots_1,
shots_2,
)
axes[1, 2].set_title("projection: 0-2")
axes[1, 2].legend().remove()
fig.tight_layout()
return fig
