class ColorbarWidget(QWidget):
colorbarChanged = Signal() # The parent should simply redraw their canvas
def __init__(self, parent=None):
super(ColorbarWidget, self).__init__(parent)
self.setWindowTitle("Colorbar")
self.setMaximumWidth(200)
self.cmap = QComboBox()
self.cmap.addItems(sorted(cm.cmap_d.keys()))
self.cmap.currentIndexChanged.connect(self.cmap_index_changed)
self.cmin = QLineEdit()
self.cmin_value = 0
self.cmin.setMaximumWidth(100)
self.cmin.editingFinished.connect(self.clim_changed)
self.cmin_layout = QHBoxLayout()
self.cmin_layout.addStretch()
self.cmin_layout.addWidget(self.cmin)
self.cmin_layout.addStretch()
self.cmax = QLineEdit()
self.cmax_value = 1
self.cmax.setMaximumWidth(100)
self.cmax.editingFinished.connect(self.clim_changed)
self.cmin.setValidator(QDoubleValidator())
self.cmax.setValidator(QDoubleValidator())
self.cmax_layout = QHBoxLayout()
self.cmax_layout.addStretch()
self.cmax_layout.addWidget(self.cmax)
self.cmax_layout.addStretch()
self.norm_layout = QHBoxLayout()
self.norm = QComboBox()
self.norm.addItems(NORM_OPTS)
self.norm.currentIndexChanged.connect(self.norm_changed)
self.powerscale = QLineEdit()
self.powerscale_value = 2
self.powerscale.setText("2")
self.powerscale.setValidator(QDoubleValidator(0.001, 100, 3))
self.powerscale.setMaximumWidth(50)
self.powerscale.editingFinished.connect(self.norm_changed)
self.powerscale.hide()
self.powerscale_label = QLabel("n=")
self.powerscale_label.hide()
self.norm_layout.addStretch()
self.norm_layout.addWidget(self.norm)
self.norm_layout.addStretch()
self.norm_layout.addWidget(self.powerscale_label)
self.norm_layout.addWidget(self.powerscale)
self.autoscale = QCheckBox("Autoscaling")
self.autoscale.setChecked(True)
self.autoscale.stateChanged.connect(self.update_clim)
self.canvas = FigureCanvas(Figure())
if parent:
# Set facecolor to match parent
self.canvas.figure.set_facecolor(
parent.palette().window().color().getRgbF())
self.ax = self.canvas.figure.add_axes([0.4, 0.05, 0.2, 0.9])
# layout
self.layout = QVBoxLayout(self)
self.layout.addWidget(self.cmap)
self.layout.addLayout(self.cmax_layout)
self.layout.addWidget(self.canvas, stretch=1)
self.layout.addLayout(self.cmin_layout)
self.layout.addLayout(self.norm_layout)
self.layout.addWidget(self.autoscale)
def set_mappable(self, mappable):
"""
When a new plot is created this method should be called with the new mappable
"""
self.ax.clear()
try: # Use current cmap
cmap = get_current_cmap(self.colorbar)
except AttributeError:
try: # else use viridis
cmap = cm.viridis
except AttributeError: # else default
cmap = None
self.colorbar = Colorbar(ax=self.ax, mappable=mappable)
self.cmin_value, self.cmax_value = mappable.get_clim()
self.update_clim_text()
self.cmap_changed(cmap)
mappable_cmap = get_current_cmap(mappable)
self.cmap.setCurrentIndex(
sorted(cm.cmap_d.keys()).index(mappable_cmap.name))
self.redraw()
def cmap_index_changed(self):
self.cmap_changed(self.cmap.currentText())
def cmap_changed(self, name):
self.colorbar.mappable.set_cmap(name)
if mpl_version_info() >= (3, 1):
self.colorbar.update_normal(self.colorbar.mappable)
else:
self.colorbar.set_cmap(name)
self.redraw()
def mappable_changed(self):
"""
Updates the colormap and min/max values of the colorbar
when the plot changes via settings.
"""
mappable_cmap = get_current_cmap(self.colorbar.mappable)
low, high = self.colorbar.mappable.get_clim()
self.cmin_value = low
self.cmax_value = high
self.update_clim_text()
self.cmap.setCurrentIndex(
sorted(cm.cmap_d.keys()).index(mappable_cmap.name))
self.redraw()
def norm_changed(self):
"""
Called when a different normalization is selected
"""
idx = self.norm.currentIndex()
if NORM_OPTS[idx] == 'Power':
self.powerscale.show()
self.powerscale_label.show()
else:
self.powerscale.hide()
self.powerscale_label.hide()
self.colorbar.mappable.set_norm(self.get_norm())
self.set_mappable(self.colorbar.mappable)
def get_norm(self):
"""
This will create a matplotlib.colors.Normalize from selected idx, limits and powerscale
"""
idx = self.norm.currentIndex()
if self.autoscale.isChecked():
cmin = cmax = None
else:
cmin = self.cmin_value
cmax = self.cmax_value
if NORM_OPTS[idx] == 'Power':
if self.powerscale.hasAcceptableInput():
self.powerscale_value = float(self.powerscale.text())
return PowerNorm(gamma=self.powerscale_value, vmin=cmin, vmax=cmax)
elif NORM_OPTS[idx] == "SymmetricLog10":
return SymLogNorm(
1e-8 if cmin is None else max(1e-8,
abs(cmin) * 1e-3),
vmin=cmin,
vmax=cmax)
else:
return Normalize(vmin=cmin, vmax=cmax)
def clim_changed(self):
"""
Called when either the min or max is changed. Will unset the autoscale.
"""
self.autoscale.blockSignals(True)
self.autoscale.setChecked(False)
self.autoscale.blockSignals(False)
self.update_clim()
def update_clim(self):
"""
This will update the clim of the plot based on min, max, and autoscale
"""
if self.autoscale.isChecked():
data = self.colorbar.mappable.get_array()
try:
try:
self.cmin_value = data[~data.mask].min()
self.cmax_value = data[~data.mask].max()
except AttributeError:
self.cmin_value = np.nanmin(data)
self.cmax_value = np.nanmax(data)
except (ValueError, RuntimeWarning):
# all values mask
pass
self.update_clim_text()
else:
if self.cmin.hasAcceptableInput():
cmin = float(self.cmin.text())
if cmin < self.cmax_value:
self.cmin_value = cmin
else: # reset values back
self.update_clim_text()
if self.cmax.hasAcceptableInput():
cmax = float(self.cmax.text())
if cmax > self.cmin_value:
self.cmax_value = cmax
else: # reset values back
self.update_clim_text()
self.colorbar.mappable.set_clim(self.cmin_value, self.cmax_value)
self.redraw()
def update_clim_text(self):
"""
Update displayed limit values based on stored ones
"""
self.cmin.setText("{:.4}".format(self.cmin_value))
self.cmax.setText("{:.4}".format(self.cmax_value))
def redraw(self):
"""
Redraws the colobar and emits signal to cause the parent to redraw
"""
self.colorbar.update_ticks()
self.colorbar.draw_all()
self.canvas.draw_idle()
self.colorbarChanged.emit()
class ColorbarWidget(QWidget):
colorbarChanged = Signal() # The parent should simply redraw their canvas
scaleNormChanged = Signal()
# register additional color maps from file
register_customized_colormaps()
# create the list
cmap_list = sorted(
[cmap for cmap in cm.cmap_d.keys() if not cmap.endswith('_r')])
def __init__(self, parent=None, default_norm_scale=None):
"""
:param default_scale: None uses linear, else either a string or tuple(string, other arguments), e.g. tuple('Power', exponent)
"""
super(ColorbarWidget, self).__init__(parent)
self.setWindowTitle("Colorbar")
self.setMaximumWidth(100)
self.cmap = QComboBox()
self.cmap.addItems(self.cmap_list)
self.cmap.currentIndexChanged.connect(self.cmap_index_changed)
self.crev = QCheckBox('Reverse')
self.crev.stateChanged.connect(self.crev_checked_changed)
self.cmin = QLineEdit()
self.cmin_value = 0
self.cmin.setMaximumWidth(100)
self.cmin.editingFinished.connect(self.clim_changed)
self.cmin_layout = QHBoxLayout()
self.cmin_layout.addStretch()
self.cmin_layout.addWidget(self.cmin)
self.cmin_layout.addStretch()
self.linear_validator = QDoubleValidator(parent=self)
self.log_validator = QDoubleValidator(MIN_LOG_VALUE,
sys.float_info.max, 3, self)
self.cmax = QLineEdit()
self.cmax_value = 1
self.cmax.setMaximumWidth(100)
self.cmax.editingFinished.connect(self.clim_changed)
self.cmax_layout = QHBoxLayout()
self.cmax_layout.addStretch()
self.cmax_layout.addWidget(self.cmax)
self.cmax_layout.addStretch()
norm_scale = 'Linear'
powerscale_value = 2
if default_norm_scale in NORM_OPTS:
norm_scale = default_norm_scale
if isinstance(default_norm_scale,
tuple) and default_norm_scale[0] in NORM_OPTS:
norm_scale = default_norm_scale[0]
if norm_scale == 'Power':
powerscale_value = float(default_norm_scale[1])
self.norm_layout = QHBoxLayout()
self.norm = QComboBox()
self.norm.addItems(NORM_OPTS)
self.norm.setCurrentText(norm_scale)
self.norm.currentIndexChanged.connect(self.norm_changed)
self.update_clim_validator()
self.powerscale = QLineEdit()
self.powerscale_value = powerscale_value
self.powerscale.setText(str(powerscale_value))
self.powerscale.setValidator(QDoubleValidator(0.001, 100, 3))
self.powerscale.setMaximumWidth(50)
self.powerscale.editingFinished.connect(self.norm_changed)
self.powerscale_label = QLabel("n=")
if norm_scale != 'Power':
self.powerscale.hide()
self.powerscale_label.hide()
self.norm_layout.addStretch()
self.norm_layout.addWidget(self.norm)
self.norm_layout.addStretch()
self.norm_layout.addWidget(self.powerscale_label)
self.norm_layout.addWidget(self.powerscale)
self.autoscale = QCheckBox("Autoscaling")
self.autoscale.setChecked(True)
self.autoscale.stateChanged.connect(self.update_clim)
self.canvas = FigureCanvas(Figure())
if parent:
# Set facecolor to match parent
self.canvas.figure.set_facecolor(
parent.palette().window().color().getRgbF())
self.ax = self.canvas.figure.add_axes([0.0, 0.02, 0.2, 0.97])
# layout
self.layout = QVBoxLayout(self)
self.layout.setContentsMargins(0, 0, 0, 0)
self.layout.setSpacing(2)
self.layout.addWidget(self.cmap)
self.layout.addWidget(self.crev)
self.layout.addLayout(self.cmax_layout)
self.layout.addWidget(self.canvas, stretch=1)
self.layout.addLayout(self.cmin_layout)
self.layout.addLayout(self.norm_layout)
self.layout.addWidget(self.autoscale)
def set_mappable(self, mappable):
"""
When a new plot is created this method should be called with the new mappable
"""
# sanity check the mappable
mappable = self._validate_mappable(mappable)
self.ax.clear()
try: # Use current cmap
cmap = get_current_cmap(self.colorbar)
except AttributeError:
# else use default
cmap = ConfigService.getString("plots.images.Colormap")
self.colorbar = Colorbar(ax=self.ax, mappable=mappable)
self.cmin_value, self.cmax_value = mappable.get_clim()
self.update_clim_text()
self.cmap_changed(cmap, False)
mappable_cmap = get_current_cmap(mappable)
if mappable_cmap.name.endswith('_r'):
self.crev.setChecked(True)
else:
self.crev.setChecked(False)
self.cmap.setCurrentIndex(
self.cmap_list.index(mappable_cmap.name.replace('_r', '')))
self.redraw()
def cmap_index_changed(self):
self.cmap_changed(self.cmap.currentText(), self.crev.isChecked())
def crev_checked_changed(self):
self.cmap_changed(self.cmap.currentText(), self.crev.isChecked())
def cmap_changed(self, name, rev):
if rev:
name += '_r'
self.colorbar.mappable.set_cmap(name)
if mpl_version_info() >= (3, 1):
self.colorbar.update_normal(self.colorbar.mappable)
else:
self.colorbar.set_cmap(name)
self.redraw()
def mappable_changed(self):
"""
Updates the colormap and min/max values of the colorbar
when the plot changes via settings.
"""
mappable_cmap = get_current_cmap(self.colorbar.mappable)
low, high = self.colorbar.mappable.get_clim()
self.cmin_value = low
self.cmax_value = high
self.update_clim_text()
self.cmap.setCurrentIndex(
sorted(cm.cmap_d.keys()).index(mappable_cmap.name))
self.redraw()
def norm_changed(self):
"""
Called when a different normalization is selected
"""
idx = self.norm.currentIndex()
if NORM_OPTS[idx] == 'Power':
self.powerscale.show()
self.powerscale_label.show()
else:
self.powerscale.hide()
self.powerscale_label.hide()
self.colorbar.mappable.set_norm(self.get_norm())
self.set_mappable(self.colorbar.mappable)
self.update_clim_validator()
self.scaleNormChanged.emit()
def get_norm(self):
"""
This will create a matplotlib.colors.Normalize from selected idx, limits and powerscale
"""
idx = self.norm.currentIndex()
if self.autoscale.isChecked():
cmin = cmax = None
else:
cmin = self.cmin_value
cmax = self.cmax_value
if NORM_OPTS[idx] == 'Power':
if self.powerscale.hasAcceptableInput():
self.powerscale_value = float(self.powerscale.text())
return PowerNorm(gamma=self.powerscale_value, vmin=cmin, vmax=cmax)
elif NORM_OPTS[idx] == "SymmetricLog10":
return SymLogNorm(
1e-8 if cmin is None else max(1e-8,
abs(cmin) * 1e-3),
vmin=cmin,
vmax=cmax)
elif NORM_OPTS[idx] == "Log":
cmin = MIN_LOG_VALUE if cmin is not None and cmin <= 0 else cmin
return LogNorm(vmin=cmin, vmax=cmax)
else:
return Normalize(vmin=cmin, vmax=cmax)
def get_colorbar_scale(self):
norm = self.colorbar.norm
scale = 'linear'
kwargs = {}
if isinstance(norm, SymLogNorm):
scale = 'symlog'
elif isinstance(norm, LogNorm):
scale = 'log'
elif isinstance(norm, PowerNorm):
scale = 'function'
kwargs = {
'functions': (lambda x: np.power(x, norm.gamma),
lambda x: np.power(x, 1 / norm.gamma))
}
return scale, kwargs
def clim_changed(self):
"""
Called when either the min or max is changed. Will unset the autoscale.
"""
self.autoscale.blockSignals(True)
self.autoscale.setChecked(False)
self.autoscale.blockSignals(False)
self.update_clim()
def update_clim(self):
"""
This will update the clim of the plot based on min, max, and autoscale
"""
if self.autoscale.isChecked():
self._autoscale_clim()
else:
self._manual_clim()
self.colorbar.mappable.set_clim(self.cmin_value, self.cmax_value)
self.redraw()
def update_clim_text(self):
"""
Update displayed limit values based on stored ones
"""
self.cmin.setText("{:.4}".format(self.cmin_value))
self.cmax.setText("{:.4}".format(self.cmax_value))
def redraw(self):
"""
Redraws the colobar and emits signal to cause the parent to redraw
"""
self.colorbar.update_ticks()
self.colorbar.draw_all()
self.canvas.draw_idle()
self.colorbarChanged.emit()
def update_clim_validator(self):
if NORM_OPTS[self.norm.currentIndex()] == "Log":
self.cmin.setValidator(self.log_validator)
self.cmax.setValidator(self.log_validator)
else:
self.cmin.setValidator(self.linear_validator)
self.cmax.setValidator(self.linear_validator)
def _autoscale_clim(self):
"""Update stored colorbar limits
The new limits are found from the colobar data """
data = self.colorbar.mappable.get_array()
norm = NORM_OPTS[self.norm.currentIndex()]
try:
try:
masked_data = data[~data.mask]
# use the smallest positive value as vmin when using log scale,
# matplotlib will take care of the data skipping part.
masked_data = masked_data[
data > 0] if norm == "Log" else masked_data
self.cmin_value = masked_data.min()
self.cmax_value = masked_data.max()
except (AttributeError, IndexError):
data = data[np.nonzero(data)] if norm == "Log" else data
self.cmin_value = np.nanmin(data)
self.cmax_value = np.nanmax(data)
except (ValueError, RuntimeWarning):
# all values mask
pass
self.update_clim_text()
def _manual_clim(self):
"""Update stored colorbar limits
The new limits are found from user input"""
if self.cmin.hasAcceptableInput():
cmin = float(self.cmin.text())
if cmin < self.cmax_value:
self.cmin_value = cmin
else: # reset values back
self.update_clim_text()
if self.cmax.hasAcceptableInput():
cmax = float(self.cmax.text())
if cmax > self.cmin_value:
self.cmax_value = cmax
else: # reset values back
self.update_clim_text()
def _create_linear_normalize_object(self):
if self.autoscale.isChecked():
cmin = cmax = None
else:
cmin = self.cmin_value
cmax = self.cmax_value
return Normalize(vmin=cmin, vmax=cmax)
def _validate_mappable(self, mappable):
"""Disable the Log option if no positive value can be found from given data (image)"""
index = NORM_OPTS.index("Log")
if mappable.get_array() is not None:
if np.all(mappable.get_array() <= 0):
self.norm.model().item(index, 0).setEnabled(False)
self.norm.setItemData(
index, "Log scale is disabled for non-positive data",
Qt.ToolTipRole)
if isinstance(mappable.norm, LogNorm):
mappable.norm = self._create_linear_normalize_object()
self.norm.blockSignals(True)
self.norm.setCurrentIndex(0)
self.norm.blockSignals(False)
else:
if not self.norm.model().item(index, 0).isEnabled():
self.norm.model().item(index, 0).setEnabled(True)
self.norm.setItemData(index, "", Qt.ToolTipRole)
return mappable
