Source code for sas.qtgui.Perspectives.Fitting.OptionsWidget
"""
Widget/logic for smearing data.
"""
import numpy as np
from PyQt5 import QtCore
from PyQt5 import QtGui
from PyQt5 import QtWidgets
from sas.qtgui.Plotting.PlotterData import Data2D
import sas.qtgui.Utilities.GuiUtils as GuiUtils
# Local UI
from sas.qtgui.Perspectives.Fitting.UI.OptionsWidgetUI import Ui_tabOptions
[docs]class DataWidgetMapper(QtWidgets.QDataWidgetMapper):
"""
Custom version of the standard QDataWidgetMapper allowing for proper
response to index change in comboboxes
"""
[docs] def addMapping(self, widget, section, propertyName=None):
if propertyName is None:
super(DataWidgetMapper, self).addMapping(widget, section)
else:
super(DataWidgetMapper, self).addMapping(widget, section, propertyName)
if isinstance(widget, QtWidgets.QComboBox):
delegate = self.itemDelegate()
widget.currentIndexChanged.connect(lambda: delegate.commitData.emit(widget))
elif isinstance(widget, QtWidgets.QCheckBox):
delegate = self.itemDelegate()
widget.stateChanged.connect(lambda: delegate.commitData.emit(widget))
[docs]class OptionsWidget(QtWidgets.QWidget, Ui_tabOptions):
plot_signal = QtCore.pyqtSignal()
QMIN_DEFAULT = 0.0005
QMAX_DEFAULT = 0.5
NPTS_DEFAULT = 150
MODEL = [
'MIN_RANGE',
'MAX_RANGE',
'NPTS',
'NPTS_FIT']
[docs] def __init__(self, parent=None, logic=None):
super(OptionsWidget, self).__init__()
self.setupUi(self)
# Logic component
self.logic = logic
self.parent = parent
# Weight radio box group
self.weightingGroup = QtWidgets.QButtonGroup()
self.weighting = 0
# Group boxes
self.boxWeighting.setEnabled(False)
self.cmdMaskEdit.setEnabled(False)
# Button groups
self.weightingGroup.addButton(self.rbWeighting1)
self.weightingGroup.addButton(self.rbWeighting2)
self.weightingGroup.addButton(self.rbWeighting3)
self.weightingGroup.addButton(self.rbWeighting4)
# Let only floats in the range edits
self.txtMinRange.setValidator(GuiUtils.DoubleValidator())
self.txtMaxRange.setValidator(GuiUtils.DoubleValidator())
# Let only ints in the number of points edit
self.txtNpts.setValidator(QtGui.QIntValidator())
# disable npts/fit - this is a read only control
self.txtNptsFit.setEnabled(False)
# Attach slots
self.cmdReset.clicked.connect(self.onRangeReset)
self.cmdMaskEdit.clicked.connect(self.onMaskEdit)
self.chkLogData.stateChanged.connect(self.toggleLogData)
# Button groups
self.weightingGroup.buttonClicked.connect(self.onWeightingChoice)
self.qmin = self.QMIN_DEFAULT
self.qmax = self.QMAX_DEFAULT
self.npts = self.NPTS_DEFAULT
self.npts_fit = self.NPTS_DEFAULT
if self.logic.data_is_loaded:
self.qmin, self.qmax, self.npts = self.logic.computeDataRange()
self.npts_fit = self.npts2fit(data=self.logic.data)
self.initModel()
self.initMapper()
self.model.blockSignals(True)
self.updateQRange(self.qmin, self.qmax, self.npts)
self.txtMaxRange.setText(str(self.qmax))
self.txtMinRange.setText(str(self.qmin))
self.txtNpts.setText(str(self.npts))
self.txtNptsFit.setText(str(self.npts_fit))
self.model.blockSignals(False)
new_font = 'font-family: -apple-system, "Helvetica Neue", "Ubuntu";'
self.label_13.setStyleSheet(new_font)
self.label_15.setStyleSheet(new_font)
[docs] def initModel(self):
"""
Initialize the state
"""
self.model = QtGui.QStandardItemModel()
for model_item in range(len(self.MODEL)):
self.model.setItem(model_item, QtGui.QStandardItem())
# Attach slot
self.model.dataChanged.connect(self.onModelChange)
[docs] def initMapper(self):
"""
Initialize model item <-> UI element mapping
"""
self.mapper = DataWidgetMapper(self)
self.mapper.setModel(self.model)
self.mapper.setOrientation(QtCore.Qt.Vertical)
self.mapper.addMapping(self.txtMinRange, self.MODEL.index('MIN_RANGE'))
self.mapper.addMapping(self.txtMaxRange, self.MODEL.index('MAX_RANGE'))
self.mapper.addMapping(self.txtNpts, self.MODEL.index('NPTS'))
self.mapper.addMapping(self.txtNptsFit, self.MODEL.index('NPTS_FIT'))
self.mapper.toFirst()
[docs] def toggleLogData(self, isChecked):
""" Toggles between log and linear data sets """
self.plot_signal.emit()
[docs] def onMaskEdit(self):
"""
Callback for running the mask editor
"""
if isinstance(self.logic.data, Data2D):
self.parent.communicate.maskEditorSignal.emit(self.logic.data)
[docs] def onRangeReset(self):
"""
Callback for resetting qmin/qmax
"""
if self.logic.data_is_loaded:
self.qmin, self.qmax, self.npts = self.logic.computeDataRange()
else:
self.qmin, self.qmax, self.npts = (self.QMIN_DEFAULT,
self.QMAX_DEFAULT,
self.NPTS_DEFAULT)
self.updateQRange(self.qmin, self.qmax, self.npts)
[docs] def onWeightingChoice(self, button):
"""
Update weighting in the fit state
"""
button_id = button.group().checkedId()
self.weighting = abs(button_id + 2)
self.plot_signal.emit()
[docs] def onModelChange(self, top, bottom):
"""
Respond to model change by updating the plot
"""
# "bottom" is unused
# update if there's something to update
item_text = self.model.item(top.row()).text()
if not item_text:
return
# Update the npts/fit value
if top.row() in [0, 1, 2]:
qmin, qmax, npts, _, _ = self.state()
# if this is a Q value, update NPt/fit
value = float(item_text)
if top.row() == 0:
qmin = value
if qmin >= qmax:
qmin = self.qmin
self.model.item(self.MODEL.index('MIN_RANGE')).setText(str(self.qmin))
self.npts_fit = self.npts2fit(data=self.logic.data, qmin=qmin, qmax=qmax, npts=npts)
elif top.row() == 1:
qmax = value
if qmax <= qmin:
qmax = self.qmax
self.model.item(self.MODEL.index('MAX_RANGE')).setText(str(self.qmax))
self.npts_fit = self.npts2fit(data=self.logic.data, qmin=qmin, qmax=qmax, npts=npts)
else:
# This is only possible for theories. Just post the number to the field, if valid
npts = int(value)
if npts <= 0:
npts = self.npts
self.npts_fit = npts
self.model.item(self.MODEL.index('NPTS')).setText(str(npts))
self.model.item(self.MODEL.index('NPTS_FIT')).setText(str(self.npts_fit))
# update the plot(s)
self.plot_signal.emit()
[docs] def setEnablementOnDataLoad(self):
"""
Enable/disable various UI elements based on data loaded
"""
is2Ddata = isinstance(self.logic.data, Data2D)
self.boxWeighting.setEnabled(True)
self.cmdMaskEdit.setEnabled(is2Ddata)
# Switch off txtNpts related controls
self.txtNpts.setEnabled(False)
self.chkLogData.setEnabled(False)
# Weighting controls
if self.logic.di_flag:
self.rbWeighting2.setEnabled(True)
self.rbWeighting2.setChecked(True)
self.onWeightingChoice(self.rbWeighting2)
else:
self.rbWeighting2.setEnabled(False)
self.rbWeighting1.setChecked(True)
self.onWeightingChoice(self.rbWeighting1)
[docs] def updateMinQ(self, q_min=None):
if q_min and (isinstance(q_min, (float, str))):
self.txtMinRange.setText(f"{float(q_min):.3}")
qmin = self.txtMinRange.text()
qmax = self.model.item(self.MODEL.index('MAX_RANGE')).text()
npts = self.model.item(self.MODEL.index('NPTS')).text()
self.updateQRange(qmin, qmax, npts)
[docs] def updateMaxQ(self, q_max=None):
if q_max and (isinstance(q_max, (float, str))):
self.txtMaxRange.setText(f"{float(q_max):.3}")
qmin = self.model.item(self.MODEL.index('MIN_RANGE')).text()
qmax = self.txtMaxRange.text()
npts = self.model.item(self.MODEL.index('NPTS')).text()
self.updateQRange(qmin, qmax, npts)
[docs] def updateQRange(self, q_range_min, q_range_max, npts):
"""
Update the local model based on calculated values
"""
qmax = str(q_range_max)
qmin = str(q_range_min)
self.model.item(self.MODEL.index('MIN_RANGE')).setText(qmin)
self.model.item(self.MODEL.index('MAX_RANGE')).setText(qmax)
self.model.item(self.MODEL.index('NPTS')).setText(str(npts))
self.qmin, self.qmax, self.npts = q_range_min, q_range_max, npts
npts_fit = self.npts2fit(self.logic.data, self.qmin, self.qmax, self.npts)
self.model.item(self.MODEL.index('NPTS_FIT')).setText(str(npts_fit))
[docs] def state(self):
"""
Returns current state of controls
"""
q_range_min = float(self.model.item(self.MODEL.index('MIN_RANGE')).text())
q_range_max = float(self.model.item(self.MODEL.index('MAX_RANGE')).text())
npts = int(self.model.item(self.MODEL.index('NPTS')).text())
npts_fit = int(self.model.item(self.MODEL.index('NPTS_FIT')).text())
log_points = self.chkLogData.isChecked()
return (q_range_min, q_range_max, npts, log_points, self.weighting)
[docs] def npts2fit(self, data=None, qmin=None, qmax=None, npts=None):
"""
return numbers of data points within qrange
:Note: This is to normalize chisq by Npts of fit
"""
npts2fit = 0
if data is None:
return npts2fit
qmin_c, qmax_c, npts_c = self.logic.computeDataRange()
if qmin is None:
qmin = qmin_c
if qmax is None:
qmax = qmax_c
if npts is None:
npts = npts_c
qmax = float(qmax)
qmin = float(qmin)
npts = int(npts)
if isinstance(data, Data2D):
radius = np.sqrt(data.qx_data * data.qx_data +
data.qy_data * data.qy_data)
index_data = (qmin <= radius) & (radius <= qmax)
index_data = (index_data) & (data.mask)
index_data = (index_data) & (np.isfinite(data.data))
npts2fit = len(data.data[index_data])
else:
for qx in data.x:
if qmax >= qx >= qmin:
npts2fit += 1
return npts2fit