class PyWeramiWindow(QtGui.QMainWindow, Ui_MainWindow):
def __init__(self, filename=None, parent=None):
super(PyWeramiWindow,self).__init__(parent)
self.settings = QtCore.QSettings("LX", "pywerami")
self.setupUi(self)
self._fig = plt.figure(facecolor="white")
self._ax = self._fig.add_subplot(111)
self._canvas = FigureCanvas(self._fig)
self._canvas.setParent(self.widget)
self._canvas.setFocusPolicy(QtCore.Qt.StrongFocus)
self.matplot.addWidget(self._canvas)
self.mpl_toolbar = NavigationToolbar(self._canvas,self.widget)
self.mpl_toolbar.hide()
self.matplot.addWidget(self.mpl_toolbar)
#set combos
self.cmaps = [m for m in plt.cm.datad if not m.endswith("_r")]
self.mapstyle.addItems(self.cmaps)
# set validators
self.levelmin.setValidator(QtGui.QDoubleValidator(self.levelmin))
self.levelmax.setValidator(QtGui.QDoubleValidator(self.levelmax))
self.levelnum.setValidator(QtGui.QIntValidator(self.levelmin))
self.levelstep.setValidator(QtGui.QDoubleValidator(self.levelstep))
self.clipmin.setValidator(QtGui.QDoubleValidator(self.clipmin))
self.clipmax.setValidator(QtGui.QDoubleValidator(self.clipmax))
# Set icons in toolbar
self.actionOpen.setIcon(QtGui.QIcon.fromTheme('document-open'))
self.actionSave.setIcon(QtGui.QIcon.fromTheme('document-save'))
self.actionSaveas.setIcon(QtGui.QIcon.fromTheme('document-save-as'))
self.actionImport.setIcon(QtGui.QIcon.fromTheme('x-office-spreadsheet'))
self.actionHome.setIcon(self.mpl_toolbar._icon('home.png'))
self.actionPan.setIcon(self.mpl_toolbar._icon('move.png'))
self.actionZoom.setIcon(self.mpl_toolbar._icon('zoom_to_rect.png'))
self.actionGrid.setIcon(QtGui.QIcon.fromTheme('format-justify-fill'))
self.actionAxes.setIcon(self.mpl_toolbar._icon('qt4_editor_options.png'))
self.actionSavefig.setIcon(self.mpl_toolbar._icon('filesave.png'))
#self.action3D.setIcon(QtGui.QIcon.fromTheme(''))
self.actionProperties.setIcon(QtGui.QIcon.fromTheme('preferences-other'))
self.actionQuit.setIcon(QtGui.QIcon.fromTheme('application-exit'))
self.actionAbout.setIcon(QtGui.QIcon.fromTheme('help-about'))
self.actionImport.triggered.connect(self.import_data)
self.actionHome.triggered.connect(self.mpl_toolbar.home)
self.actionPan.triggered.connect(self.plotpan)
self.actionZoom.triggered.connect(self.plotzoom)
self.actionGrid.triggered.connect(self.plotgrid)
self.actionAxes.triggered.connect(self.mpl_toolbar.edit_parameters)
self.actionSavefig.triggered.connect(self.mpl_toolbar.save_figure)
self.actionProperties.triggered.connect(self.edit_options)
self.actionQuit.triggered.connect(self.close)
if filename:
self.import_data(filename)
# ready
self.statusbar.showMessage("Ready", 5000)
def closeEvent(self,event):
reply=QtGui.QMessageBox.question(self,'Message',"Are you sure to quit?",QtGui.QMessageBox.Yes,QtGui.QMessageBox.No)
if reply==QtGui.QMessageBox.Yes:
event.accept()
else:
event.ignore()
def import_data(self, filename=None):
if not filename:
filename = QtGui.QFileDialog.getOpenFileName(self, "Import tab file", ".", "TAB (*.tab);;All files (*.*)")
if filename:
self.data = WeramiData.from_tab(filename)
# populate listview and setup properties
self.props = {}
self._model = QtGui.QStandardItemModel(self.listView)
for var in self.data.dep:
item = QtGui.QStandardItem(var)
item.setCheckable(True)
self._model.appendRow(item)
self.default_var_props(var)
self.listView.setModel(self._model)
self.listView.show()
# connect listview signals
self.varSel = self.listView.selectionModel()
self.varSel.selectionChanged.connect(self.on_var_changed)
self._model.itemChanged.connect(self.plot)
# buttons signals
self.buttonBox.button(QtGui.QDialogButtonBox.Apply).clicked.connect(self.apply_props)
self.buttonBox.button(QtGui.QDialogButtonBox.RestoreDefaults).clicked.connect(self.restore_props)
self.contcolor.clicked.connect(self.contours_color)
self.action3D.triggered.connect(self.switch3d)
# signals to calculate step size
self.levelmin.editingFinished.connect(self.step_from_levels)
self.levelmax.editingFinished.connect(self.step_from_levels)
self.levelnum.editingFinished.connect(self.step_from_levels)
self.setlevels.toggled.connect(self.step_from_levels)
# all done focus
self.action3D.setChecked(False) # no 3d on import
self.varSel.setCurrentIndex(self._model.index(0, 0), QtGui.QItemSelectionModel.ClearAndSelect | QtGui.QItemSelectionModel.Rows)
self.listView.setFocus()
self.plot()
self.statusbar.showMessage("Data from {} imported".format(self.data.label), 5000)
def contours_color(self):
col = QtGui.QColorDialog.getColor()
if col.isValid():
self.contcolor.setStyleSheet("background-color: {}".format(col.name()))
def step_from_levels(self):
if self.setlevels.isChecked():
step = (float(self.levelmax.text()) - float(self.levelmin.text())) / (int(self.levelnum.text()) - 1)
self.levelstep.setText(repr(step))
self.props[self.var]['step'] = step
def default_var_props(self, var):
data = self.data.get_var(var)
prop = {}
#levels
prop['min'] = data.min()
prop['max'] = data.max()
prop['num'] = 10
prop['step'] = (prop['max'] - prop['min']) / (prop['num'] - 1)
prop['levels'] = 'num'
prop['type'] = 'linear'
#style
prop['fill'] = False
prop['opacity'] = 100
prop['cmap'] = 'jet'
prop['contours'] = 'color'
prop['color'] = '#000000'
prop['label'] = False
#processing
prop['resample'] = 1
prop['median'] = 1
prop['gauss'] = 0
prop['clipmin'] = data.min()
prop['clipmax'] = data.max()
self.props[var] = prop
def set_var_props(self, var):
#levels
self.levelmin.setText(repr(self.props[var]['min']))
self.levelmax.setText(repr(self.props[var]['max']))
self.levelnum.setText(repr(self.props[var]['num']))
self.levelstep.setText(repr(self.props[var]['step']))
if self.props[var]['levels'] == 'num':
self.setlevels.setChecked(True)
else:
self.setstep.setChecked(True)
if self.props[var]['type'] == 'linear':
self.linlevel.setChecked(True)
else:
self.cdflevel.setChecked(True)
#style
if self.props[var]['fill']:
self.fillstyle.setChecked(True)
else:
self.fillstyle.setChecked(False)
self.opacity.setValue(self.props[var]['opacity'])
self.mapstyle.setCurrentIndex(self.cmaps.index(self.props[var]['cmap']))
self.contcolor.setStyleSheet("background-color: {}".format(self.props[var]['color']))
if self.props[var]['contours'] == 'map':
self.contcheckmap.setChecked(True)
elif self.props[var]['contours'] == 'color':
self.contcheckcolor.setChecked(True)
else:
self.contchecknone.setChecked(True)
if self.props[var]['label']:
self.contlabel.setChecked(True)
else:
self.contlabel.setChecked(False)
#processing
self.resample.setValue(self.props[var]['resample'])
self.filtersize.setValue(self.props[var]['median'])
self.filtersigma.setValue(self.props[var]['gauss'])
self.clipmin.setText(repr(self.props[var]['clipmin']))
self.clipmax.setText(repr(self.props[var]['clipmax']))
def on_var_changed(self, selected):
self.var = self.data.dep[selected.indexes()[0].row()]
self.set_var_props(self.var)
if self.action3D.isChecked():
self.plot()
def apply_props(self):
#levels
self.props[self.var]['min'] = float(self.levelmin.text())
self.props[self.var]['max'] = float(self.levelmax.text())
self.props[self.var]['num'] = int(self.levelnum.text())
self.props[self.var]['step'] = float(self.levelstep.text())
if self.setlevels.isChecked():
self.props[self.var]['levels'] = 'num'
else:
self.props[self.var]['levels'] = 'step'
if self.linlevel.isChecked():
self.props[self.var]['type'] = 'linear'
else:
self.props[self.var]['type'] = 'cdf'
#style
if self.fillstyle.isChecked():
self.props[self.var]['fill'] = True
else:
self.props[self.var]['fill'] = False
self.props[self.var]['opacity'] = self.opacity.value()
self.props[self.var]['cmap'] = str(self.mapstyle.currentText())
self.props[self.var]['color'] = str(self.contcolor.palette().color(1).name())
if self.contcheckmap.isChecked():
self.props[self.var]['contours'] = 'map'
elif self.contcheckcolor.isChecked():
self.props[self.var]['contours'] = 'color'
else:
self.props[self.var]['contours'] = ''
if self.contlabel.isChecked():
self.props[self.var]['label'] = True
else:
self.props[self.var]['label'] = False
#processing
self.props[self.var]['resample'] = self.resample.value()
self.props[self.var]['median'] = self.filtersize.value()
self.props[self.var]['gauss'] = self.filtersigma.value()
self.props[self.var]['clipmin'] = float(self.clipmin.text())
self.props[self.var]['clipmax'] = float(self.clipmax.text())
self.plot()
def restore_props(self):
self.default_var_props(self.var)
self.set_var_props(self.var)
self.plot()
def edit_options(self):
dlg = OptionsForm(self)
dlg.exec_()
def plotpan(self):
self.actionZoom.setChecked(False)
self.mpl_toolbar.pan()
def plotzoom(self):
self.actionPan.setChecked(False)
self.mpl_toolbar.zoom()
def plotgrid(self):
plt.grid()
self._canvas.draw()
def switch3d(self):
if not self.action3D.isChecked():
self._fig.clear()
self._ax = self._fig.add_subplot(111)
else:
self._fig.clear()
self._ax = self._fig.add_subplot(111, projection='3d')
self.plot()
def plot(self):
self._ax.cla()
if not self.action3D.isChecked():
extent = self.data.get_extent()
i = 0
while self._model.item(i):
if self._model.item(i).checkState():
CS = None
var = str(self._model.item(i).text())
# get data, smooth and clip
data = self.data.get_var(var,nan=np.float(self.settings.value("nan", "NaN", type=str)))
if self.props[var]['resample'] > 1:
data = np.ma.array(ndimage.zoom(data.filled(0), self.props[var]['resample']), mask=ndimage.zoom(data.mask, self.props[var]['resample'], order=0))
if self.props[var]['median'] > 1:
data = np.ma.array(ndimage.median_filter(data, size=self.props[var]['median']*self.props[var]['resample']), mask=data.mask)
if self.props[var]['gauss'] > 0:
data = np.ma.array(ndimage.gaussian_filter(data, sigma=self.props[var]['gauss']*self.props[var]['resample']), mask=data.mask)
data = np.ma.masked_outside(data, self.props[var]['clipmin'], self.props[var]['clipmax'])
if self.props[var]['fill']:
self._ax.imshow(data, interpolation='none', origin='lower', extent=extent, aspect='auto', cmap=plt.get_cmap(self.props[var]['cmap']), alpha=self.props[var]['opacity']/100.0)
if self.props[var]['type'] == 'linear':
if self.props[var]['levels'] == 'num':
clevels = np.linspace(self.props[var]['min'], self.props[var]['max'], self.props[var]['num'])
else:
# trick to include max in levels
clevels = np.arange(self.props[var]['min'], self.props[var]['max'] + 10**np.ceil(np.log10(np.abs(self.props[var]['max']))), self.props[var]['step'])
else:
# cdf based on histogram binned acording to the Freedman-Diaconis rule
v = np.sort(data.compressed())
IQR = v[int(round((v.size-1) * float(0.75)))] - v[int(round((v.size-1) * float(0.25)))]
bin_size = 2 * IQR * v.size**(-1.0/3)
nbins = int(round(max(self.props[var]['num'], (v[-1]-v[0]) / (bin_size+0.001))))
hist, bin_edges = np.histogram(v, bins=nbins)
cdf = np.cumsum(hist)
cdfx = np.cumsum(np.diff(bin_edges)) + bin_edges[:2].sum()/2
clevels = np.interp(np.linspace(cdf[0],cdf[-1],self.props[var]['num'] + 2)[1:-1], cdf, cdfx)
if self.props[var]['contours'] == 'map':
CS = self._ax.contour(self.data.get_xrange(self.props[var]['resample']), self.data.get_yrange(self.props[var]['resample']), data, clevels, cmap=plt.get_cmap(self.props[var]['cmap']))
elif self.props[var]['contours'] == 'color':
CS = self._ax.contour(self.data.get_xrange(self.props[var]['resample']), self.data.get_yrange(self.props[var]['resample']), data, clevels, colors=self.props[var]['color'])
if self.props[var]['label'] and CS:
self._ax.clabel(CS, fontsize=8, inline=1)
i += 1
self._ax.axis(extent)
plt.title(self.data.label)
else:
# get data, smooth and clip
data = self.data.get_var(self.var)
if self.props[self.var]['resample'] > 1:
data = np.ma.array(ndimage.zoom(data.filled(0), self.props[self.var]['resample']), mask=ndimage.zoom(data.mask, self.props[self.var]['resample'], order=0))
if self.props[self.var]['median'] > 1:
data = np.ma.array(ndimage.median_filter(data, size=self.props[self.var]['median']*self.props[self.var]['resample']), mask=data.mask)
if self.props[self.var]['gauss'] > 0:
data = np.ma.array(ndimage.gaussian_filter(data, sigma=self.props[self.var]['gauss']*self.props[self.var]['resample']), mask=data.mask)
data = np.ma.masked_outside(data, self.props[self.var]['clipmin'], self.props[self.var]['clipmax'])
x,y = np.meshgrid(self.data.get_xrange(self.props[self.var]['resample']), self.data.get_yrange(self.props[self.var]['resample']))
self._ax.plot_surface(x, y, data.filled(np.NaN), vmin=data.min(), vmax=data.max(), cmap=plt.get_cmap(self.props[self.var]['cmap']), linewidth=0.5, alpha=self.props[self.var]['opacity']/100.0)
self._ax.view_init(azim=235, elev=30)
plt.xlabel(self.data.ind[self.data.xvar]['name'])
plt.ylabel(self.data.ind[self.data.yvar]['name'])
plt.tight_layout()
self._canvas.draw()
def _icon(self, name):
if name.startswith(os.path.sep):
return QtGui.QIcon(name)
return NavigationToolbar._icon(self, name)
