Merge branch 'github-master'

release notes:
==============
Features:
- centralize all functionalities of PyLoT and control them from within the main GUI
- handling multiple events inside GUI with project files (save and load work progress)
- GUI based adjustments of pick parameters and I/O
- interactive tuning of parameters from within the GUI
- call automatic picking algorithm from within the GUI
- comparison of automatic with manual picks for multiple events using clear differentiation of manual picks into 'tune' and 'test-set' (beta)
- manual picking of different (user defined) phase types
- phase onset estimation with ObsPy TauPy

- interactive zoom/scale functionalities in all plots (mousewheel, pan, pan-zoom)
- array map to visualize stations and control onsets (beta feature, switch to manual picks not implemented)

Platform support:
- python 3 support
- Windows support

Performance:
- multiprocessing for automatic picking and restitution of multiple stations
- use pyqtgraph library for better performance on main waveform plot

Visualization:
- pick uncertainty (quality classes) visualization with gradients
- pick color unification for all plots
- new icons and stylesheets

Known Issues:

PyLoT.py

@@ -1633,7 +1633,7 @@ class MainWindow(QMainWindow):
         if self.pg:
             self.finish_pg_plot()
         else:
-            self._max_xlims = self.dataPlot.getXLims()
+            self._max_xlims = self.dataPlot.getXLims(self.dataPlot.axes[0])
         plotWidget = self.getPlotWidget()
         plotDict = plotWidget.getPlotDict()
         pos = plotDict.keys()
@@ -1707,7 +1707,7 @@ class MainWindow(QMainWindow):
             self.dataPlot.plotWidget.hideAxis('bottom')
             self.dataPlot.plotWidget.hideAxis('left')
         else:
-            self.dataPlot.getAxes().cla()
+            self.dataPlot.axes[0].cla()
         self.loadlocationaction.setEnabled(False)
         self.auto_tune.setEnabled(False)
         self.auto_pick.setEnabled(False)
@@ -2314,7 +2314,7 @@ class MainWindow(QMainWindow):
         if self.pg:
             pw = self.getPlotWidget().plotWidget
         else:
-            ax = self.getPlotWidget().axes
+            ax = self.getPlotWidget().axes[0]
         ylims = np.array([-.5, +.5]) + plotID
 
         stat_picks = self.getPicks(type=picktype)[station]
@@ -2572,7 +2572,7 @@ class MainWindow(QMainWindow):
                 lon = event.origins[0].longitude
                 self.array_map.eventLoc = (lat, lon)
         if self.get_current_event():
-            self.array_map.refresh_drawings(self.get_current_event().getPicks())
+            self.array_map.refresh_drawings(self.get_current_event().getAutopicks())
         self._eventChanged[1] = False
 
     def init_event_table(self, tabindex=2):

README.md

@@ -66,27 +66,41 @@ PyLoT has been tested on Mac OSX (10.11), Debian Linux 8 and on Windows 10.
 
 #### Features:
 
-- event organisation in project files and waveform visualisation
-- consistent manual phase picking through predefined SNR dependant zoom level
-- consistent automatic phase picking routines using Higher Order Statistics, AIC and Autoregression
-- interactive tuning of auto-pick parameters
-- uniform uncertainty estimation from waveform's properties for automatic and manual picks
-- pdf representation and comparison of picks taking the uncertainty intrinsically into account 
-- Richter and moment magnitude estimation
-- location determination with external installation of [NonLinLoc](http://alomax.free.fr/nlloc/index.html)
+- centralize all functionalities of PyLoT and control them from within the main GUI
+- handling multiple events inside GUI with project files (save and load work progress)
+- GUI based adjustments of pick parameters and I/O
+- interactive tuning of parameters from within the GUI
+- call automatic picking algorithm from within the GUI
+- comparison of automatic with manual picks for multiple events using clear differentiation of manual picks into 'tune' and 'test-set' (beta)
+- manual picking of different (user defined) phase types
+- phase onset estimation with ObsPy TauPy
+- interactive zoom/scale functionalities in all plots (mousewheel, pan, pan-zoom)
+- array map to visualize stations and control onsets (beta feature, switch to manual picks not implemented)
 
-#### Known issues:
+##### Platform support:
+- Python 3 support
+- Windows support
 
-- Sometimes an error might occur when using Qt
+##### Performance:
+- multiprocessing for automatic picking and restitution of multiple stations
+- use pyqtgraph library for better performance on main waveform plot
 
-We hope to solve these with the next release.
+##### Visualization:
+- pick uncertainty (quality classes) visualization with gradients
+- pick color unification for all plots
+- new icons and stylesheets
+
+#### Known Issues:
+- some Qt related errors might occur at runtime
+- filter toggle not working in pickDlg
+- PyLoT data structure requires at least three parent directories for waveform data directory
 
 ## Staff
 
 Original author(s): L. Kueperkoch, S. Wehling-Benatelli, M. Bischoff (PILOT)
 
-Developer(s): S. Wehling-Benatelli, L. Kueperkoch, K. Olbert, M. Bischoff,
-              C. Wollin, M. Rische, M. Paffrath
+Developer(s): S. Wehling-Benatelli, L. Kueperkoch, M. Paffrath, K. Olbert,
+M. Bischoff, C. Wollin, M. Rische
 
 Others: A. Bruestle, T. Meier, W. Friederich
 

pylot/core/util/widgets.py

@@ -624,6 +624,7 @@ class PylotCanvas(FigureCanvas):
         # initialize super class
         super(PylotCanvas, self).__init__(self.figure)
         self.setParent(parent)
+        self.orig_parent = parent
 
         if multicursor:
             # add a cursor for station selection
@@ -917,7 +918,7 @@ class PylotCanvas(FigureCanvas):
         nsc.sort()
         nsc.reverse()
 
-        style = self.parent()._style
+        style = self.orig_parent._style
         linecolor = style['linecolor']['rgba_mpl']
 
         for n, (network, station, channel) in enumerate(nsc):