def adjust_layout(self,
event: Optional[Event] = None, *,
refresh_canvas: bool = False) -> None:
"""Optimize subplot grid and buttons width for best fit, then adjust
layout based on window size.
:param event: Event spent by figure `mpl_connect` 'resize_event'.
:param refresh_canvas: Force redrawing figure canvas.
"""
# Compute figure area for later export
bbox = Bbox([[0.0, 0.065], [1.0, 1.0]])
bbox_pixels = bbox.transformed(self.figure.transFigure)
self.bbox_inches = bbox_pixels.transformed(
self.figure.dpi_scale_trans.inverted())
# Refresh button size, in case the number of tabs has changed
buttons_width = 1.0 / (len(self.tabs_data) + 1)
for i, tab in enumerate(self.tabs_data.values()):
tab["button_axcut"].set_position(
[buttons_width * (i + 0.5), 0.01, buttons_width, 0.05])
# Re-arrange subplots in case figure aspect ratio has changed
axes = self.tab_active["axes"]
num_subplots = len(axes)
figure_extent = self.figure.get_window_extent()
figure_ratio = figure_extent.width / figure_extent.height
num_rows_1 = max(1, floor(sqrt(num_subplots / figure_ratio)))
num_cols_1 = ceil(num_subplots / num_rows_1)
num_cols_2 = ceil(sqrt(num_subplots * figure_ratio))
num_rows_2 = ceil(num_subplots / num_cols_2)
if num_rows_1 * num_cols_1 < num_rows_2 * num_cols_2:
num_rows, num_cols = map(int, (num_rows_1, num_cols_1))
else:
num_rows, num_cols = map(int, (num_rows_2, num_cols_2))
for i, ax in enumerate(axes, 1):
ax.change_geometry(num_rows, num_cols, i)
# Adjust layout: namely margins between subplots
right_margin = 0.03
if self.legend is not None:
legend_extent = self.legend.get_window_extent()
legend_width_rel = legend_extent.transformed(
self.figure.transFigure.inverted()).width
right_margin += legend_width_rel
self.figure.subplots_adjust(
bottom=0.10, top=0.92, left=0.04, right=1.0-right_margin,
wspace=0.40, hspace=0.30)
# Refresh figure canvas if requested
if refresh_canvas:
self.refresh()
vert = tag_path.vertices - dims / 2.
mult = tagline_scale_fudge * (plot_dims / dims).min()
mult = [mult, -mult] # y axis is inverted
offset = mne_corners[-1] - np.array([mne_clip.get_extents().size[0] / 2.,
-dims[1]]) - tagline_offset_fudge
tag_clip = Path(offset + vert * mult, tag_path.codes)
tag_patch = PathPatch(tag_clip, facecolor='k', edgecolor='none', zorder=10)
ax.add_patch(tag_patch)
yl = ax.get_ylim()
yy = np.max([tag_clip.vertices.max(0)[-1],
tag_clip.vertices.min(0)[-1]])
ax.set_ylim(np.ceil(yy), yl[-1])
# only save actual image extent plus a bit of padding
extent = Bbox(np.c_[ax.get_xlim(), ax.get_ylim()])
extent = extent.transformed(ax.transData + fig.dpi_scale_trans.inverted())
plt.draw()
plt.savefig(op.join(static_dir, 'mne_logo.png'),
bbox_inches=extent.expanded(1.2, 1.))
plt.close()
# 92x22 image
w_px = 92
h_px = 22
center_fudge = np.array([12, 0.5])
scale_fudge = 2.1
rcParams.update({'font.sans-serif': ['Primetime'], 'font.weight': 'black'})
x = np.linspace(-8., 8., w_px / 2.)
y = np.linspace(-3., 3., h_px / 2.)
X, Y = np.meshgrid(x, y)
# initialize figure (no axes, margins, etc)
def event_plot(events, rowHeight=100):
genes = numpy.unique(
numpy.char.partition(events.rownames.astype(str), "_")[:, 0])
x = pandas.DataFrame(numpy.zeros((len(genes), events.shape[1]), dtype=int),
index=numpy.array(genes),
columns=events.colnames)
fig = plt.gcf()
bbox = Bbox([[0.8 * fig.dpi, 0.5 * fig.dpi],
[(fig.get_figwidth() - 0.5) * fig.dpi,
0.3 * len(x) * fig.dpi]])
plt.axes(bbox.transformed(plt.gcf().transFigure.inverted()).bounds)
for feature in events.rownames:
gene, eventType = feature.split("_", 1)
i = genes.searchsorted(gene)
if eventType == "mut":
shift = 0
elif eventType == "gain":
shift = 1
elif eventType == "loss":
shift = 2
x.values[i] |= events[[feature]].events[0].astype(int) << shift
geneOrder = (x.values > 0).sum(1).argsort()
sampleOrder = numpy.lexsort(x.values[geneOrder])[::-1]
colours = matplotlib.colors.ListedColormap(
["#dddddd", "#377eb8", "#2ca25f", "#756bb1", "#de2d26", "#ff7f00"])
plt.imshow(x.values[geneOrder][:, sampleOrder],
aspect="auto",
interpolation="none",
cmap=colours,
origin="lower",
vmin=0,
vmax=5)
plt.grid(ls="-", c="white", axis="x")
plt.setp(plt.gca().spines.values(), color="#888888", alpha=0)
for tic in plt.gca().xaxis.get_major_ticks():
tic.tick1On = tic.tick2On = False
for tic in plt.gca().yaxis.get_major_ticks():
tic.tick1On = tic.tick2On = False
try:
plt.yticks(numpy.arange(len(x)), x.index[geneOrder])
except AttributeError:
pass
plt.gca().yaxis.set_minor_locator(
matplotlib.ticker.FixedLocator(
numpy.linspace(-0.5,
len(x) - 0.5, x.shape[0] + 1)))
plt.grid(ls="-", c="white", axis="y", lw=4, alpha=1, which="minor")
plt.grid(ls="None", axis="y", which="major")
for tic in plt.gca().yaxis.get_minor_ticks():
tic.tick1On = tic.tick2On = False
plt.gcf().patch.set_facecolor("white")
plt.xlim(0, x.ix[genes].any(0).sum())
