Vehicle-Anti-Theft-Face-Rec.../venv/Lib/site-packages/matplotlib/tests/test_backend_bases.py

159 lines
5.8 KiB
Python

import re
from matplotlib.backend_bases import (
FigureCanvasBase, LocationEvent, MouseButton, MouseEvent,
NavigationToolbar2, RendererBase)
import matplotlib.pyplot as plt
import matplotlib.transforms as transforms
import matplotlib.path as path
import os
import numpy as np
import pytest
def test_uses_per_path():
id = transforms.Affine2D()
paths = [path.Path.unit_regular_polygon(i) for i in range(3, 7)]
tforms_matrices = [id.rotate(i).get_matrix().copy() for i in range(1, 5)]
offsets = np.arange(20).reshape((10, 2))
facecolors = ['red', 'green']
edgecolors = ['red', 'green']
def check(master_transform, paths, all_transforms,
offsets, facecolors, edgecolors):
rb = RendererBase()
raw_paths = list(rb._iter_collection_raw_paths(
master_transform, paths, all_transforms))
gc = rb.new_gc()
ids = [path_id for xo, yo, path_id, gc0, rgbFace in
rb._iter_collection(
gc, master_transform, all_transforms,
range(len(raw_paths)), offsets,
transforms.AffineDeltaTransform(master_transform),
facecolors, edgecolors, [], [], [False],
[], 'screen')]
uses = rb._iter_collection_uses_per_path(
paths, all_transforms, offsets, facecolors, edgecolors)
if raw_paths:
seen = np.bincount(ids, minlength=len(raw_paths))
assert set(seen).issubset([uses - 1, uses])
check(id, paths, tforms_matrices, offsets, facecolors, edgecolors)
check(id, paths[0:1], tforms_matrices, offsets, facecolors, edgecolors)
check(id, [], tforms_matrices, offsets, facecolors, edgecolors)
check(id, paths, tforms_matrices[0:1], offsets, facecolors, edgecolors)
check(id, paths, [], offsets, facecolors, edgecolors)
for n in range(0, offsets.shape[0]):
check(id, paths, tforms_matrices, offsets[0:n, :],
facecolors, edgecolors)
check(id, paths, tforms_matrices, offsets, [], edgecolors)
check(id, paths, tforms_matrices, offsets, facecolors, [])
check(id, paths, tforms_matrices, offsets, [], [])
check(id, paths, tforms_matrices, offsets, facecolors[0:1], edgecolors)
def test_get_default_filename(tmpdir):
plt.rcParams['savefig.directory'] = str(tmpdir)
fig = plt.figure()
canvas = FigureCanvasBase(fig)
filename = canvas.get_default_filename()
assert filename == 'image.png'
def test_canvas_change():
fig = plt.figure()
# Replaces fig.canvas
canvas = FigureCanvasBase(fig)
# Should still work.
plt.close(fig)
assert not plt.fignum_exists(fig.number)
@pytest.mark.backend('pdf')
def test_non_gui_warning(monkeypatch):
plt.subplots()
monkeypatch.setitem(os.environ, "DISPLAY", ":999")
with pytest.warns(UserWarning) as rec:
plt.show()
assert len(rec) == 1
assert ('Matplotlib is currently using pdf, which is a non-GUI backend'
in str(rec[0].message))
with pytest.warns(UserWarning) as rec:
plt.gcf().show()
assert len(rec) == 1
assert ('Matplotlib is currently using pdf, which is a non-GUI backend'
in str(rec[0].message))
@pytest.mark.parametrize(
"x, y", [(42, 24), (None, 42), (None, None), (200, 100.01), (205.75, 2.0)])
def test_location_event_position(x, y):
# LocationEvent should cast its x and y arguments to int unless it is None.
fig, ax = plt.subplots()
canvas = FigureCanvasBase(fig)
event = LocationEvent("test_event", canvas, x, y)
if x is None:
assert event.x is None
else:
assert event.x == int(x)
assert isinstance(event.x, int)
if y is None:
assert event.y is None
else:
assert event.y == int(y)
assert isinstance(event.y, int)
if x is not None and y is not None:
assert re.match(
"x={} +y={}".format(ax.format_xdata(x), ax.format_ydata(y)),
ax.format_coord(x, y))
ax.fmt_xdata = ax.fmt_ydata = lambda x: "foo"
assert re.match("x=foo +y=foo", ax.format_coord(x, y))
def test_interactive_zoom():
fig, ax = plt.subplots()
ax.set(xscale="logit")
assert ax.get_navigate_mode() is None
tb = NavigationToolbar2(fig.canvas)
tb.zoom()
assert ax.get_navigate_mode() == 'ZOOM'
xlim0 = ax.get_xlim()
ylim0 = ax.get_ylim()
# Zoom from x=1e-6, y=0.1 to x=1-1e-5, 0.8 (data coordinates, "d").
d0 = (1e-6, 0.1)
d1 = (1-1e-5, 0.8)
# Convert to screen coordinates ("s"). Events are defined only with pixel
# precision, so round the pixel values, and below, check against the
# corresponding xdata/ydata, which are close but not equal to d0/d1.
s0 = ax.transData.transform(d0).astype(int)
s1 = ax.transData.transform(d1).astype(int)
# Zoom in.
start_event = MouseEvent(
"button_press_event", fig.canvas, *s0, MouseButton.LEFT)
fig.canvas.callbacks.process(start_event.name, start_event)
stop_event = MouseEvent(
"button_release_event", fig.canvas, *s1, MouseButton.LEFT)
fig.canvas.callbacks.process(stop_event.name, stop_event)
assert ax.get_xlim() == (start_event.xdata, stop_event.xdata)
assert ax.get_ylim() == (start_event.ydata, stop_event.ydata)
# Zoom out.
start_event = MouseEvent(
"button_press_event", fig.canvas, *s1, MouseButton.RIGHT)
fig.canvas.callbacks.process(start_event.name, start_event)
stop_event = MouseEvent(
"button_release_event", fig.canvas, *s0, MouseButton.RIGHT)
fig.canvas.callbacks.process(stop_event.name, stop_event)
# Absolute tolerance much less than original xmin (1e-7).
assert ax.get_xlim() == pytest.approx(xlim0, rel=0, abs=1e-10)
assert ax.get_ylim() == pytest.approx(ylim0, rel=0, abs=1e-10)
tb.zoom()
assert ax.get_navigate_mode() is None