#
-# Copyright 2008 Free Software Foundation, Inc.
+# Copyright 2008, 2009 Free Software Foundation, Inc.
#
# This file is part of GNU Radio
#
# Boston, MA 02110-1301, USA.
#
+##################################################
+# conditional disconnections of wx flow graph
+##################################################
+import wx
+from gnuradio import gr
+
+class wxgui_hb(object):
+ """
+ The wxgui hier block helper/wrapper class:
+ A hier block should inherit from this class to make use of the wxgui connect method.
+ To use, call wxgui_connect in place of regular connect; self.win must be defined.
+ The implementation will conditionally enable the copy block after the source (self).
+ This condition depends on weather or not the window is visible with the parent notebooks.
+ This condition will be re-checked on every ui update event.
+ """
+
+ def wxgui_connect(self, *points):
+ """
+ Use wxgui connect when the first point is the self source of the hb.
+ The win property of this object should be set to the wx window.
+ When this method tries to connect self to the next point,
+ it will conditionally make this connection based on the visibility state.
+ All other points will be connected normally.
+ """
+ try:
+ assert points[0] == self or points[0][0] == self
+ copy = gr.copy(self._hb.input_signature().sizeof_stream_item(0))
+ handler = self._handler_factory(copy.set_enabled)
+ handler(False) #initially disable the copy block
+ self._bind_to_visible_event(win=self.win, handler=handler)
+ points = list(points)
+ points.insert(1, copy) #insert the copy block into the chain
+ except (AssertionError, IndexError): pass
+ self.connect(*points) #actually connect the blocks
+
+ @staticmethod
+ def _handler_factory(handler):
+ """
+ Create a function that will cache the visibility flag,
+ and only call the handler when that flag changes.
+ @param handler the function to call on a change
+ @return a function of 1 argument
+ """
+ cache = [None]
+ def callback(visible):
+ if cache[0] == visible: return
+ cache[0] = visible
+ #print visible, handler
+ handler(visible)
+ return callback
+
+ @staticmethod
+ def _bind_to_visible_event(win, handler):
+ """
+ Bind a handler to a window when its visibility changes.
+ Specifically, call the handler when the window visibility changes.
+ This condition is checked on every update ui event.
+ @param win the wx window
+ @param handler a function of 1 param
+ """
+ #is the window visible in the hierarchy
+ def is_wx_window_visible(my_win):
+ while True:
+ parent = my_win.GetParent()
+ if not parent: return True #reached the top of the hierarchy
+ #if we are hidden, then finish, otherwise keep traversing up
+ if isinstance(parent, wx.Notebook) and parent.GetCurrentPage() != my_win: return False
+ my_win = parent
+ #call the handler, the arg is shown or not
+ def handler_factory(my_win, my_handler):
+ def callback(evt):
+ my_handler(is_wx_window_visible(my_win))
+ evt.Skip() #skip so all bound handlers are called
+ return callback
+ handler = handler_factory(win, handler)
+ #bind the handler to all the parent notebooks
+ win.Bind(wx.EVT_UPDATE_UI, handler)
+
+##################################################
+# Helpful Functions
+##################################################
+
#A macro to apply an index to a key
index_key = lambda key, i: "%s_%d"%(key, i+1)
@param samples the array of real values
@return a tuple of min, max
"""
- scale_factor = 3
+ factor = 2.0
mean = numpy.average(samples)
- rms = numpy.max([scale_factor*((numpy.sum((samples-mean)**2)/len(samples))**.5), .1])
- min = mean - rms
- max = mean + rms
- return min, max
+ std = numpy.std(samples)
+ fft = numpy.abs(numpy.fft.fft(samples - mean))
+ envelope = 2*numpy.max(fft)/len(samples)
+ ampl = max(std, envelope) or 0.1
+ return mean - factor*ampl, mean + factor*ampl
+
+def get_min_max_fft(fft_samps):
+ """
+ Get the minimum and maximum bounds for an array of fft samples.
+ @param samples the array of real values
+ @return a tuple of min, max
+ """
+ #get the peak level (max of the samples)
+ peak_level = numpy.max(fft_samps)
+ #separate noise samples
+ noise_samps = numpy.sort(fft_samps)[:len(fft_samps)/2]
+ #get the noise floor
+ noise_floor = numpy.average(noise_samps)
+ #get the noise deviation
+ noise_dev = numpy.std(noise_samps)
+ #determine the maximum and minimum levels
+ max_level = peak_level
+ min_level = noise_floor - abs(2*noise_dev)
+ return min_level, max_level
projects / debian / gnuradio / blobdiff