build.rsudp.c_plot.Plot.run() - Code Metrics - Inspection of "adding plot test" - raspishake/rsudp - Measure and Improve Code Quality continuously with Scrutinizer

Test Failed
Push — coverage ( 0196fb...b39e1a )

by Ian
created 2021-10-17 21:35 UTC
build.rsudp.c_plot.Plot.run() C

↳ Parent: build.rsudp.c_plot
Complexity

Conditions
Size

Total Lines	40
Code Lines	30
Duplication

Lines	0
Ratio	0 %
Importance

Changes
Metric	Value
eloc	30
dl	0
loc	40
rs	6.6666
c	0
b	0
f	0
cc	9
nop	1
import os, sys, platform
import pkg_resources as pr
import time
import math
import numpy as np
from datetime import datetime, timedelta
import rsudp.raspberryshake as rs
from rsudp import printM, printW, printE, helpers, scap_dir
from rsudp.test import TEST
import linecache
sender = 'plot.py'
QT = False
QtGui = False
PhotoImage = False
try:		# test for matplotlib and exit if import fails
	from matplotlib import use
	try:	# no way to know what machines can handle what software, but Tk is more universal
		use('Qt5Agg')	# try for Qt because it's better and has less threatening errors
		from PyQt5 import QtGui
		QT = True
	except Exception as e:
		printW('Qt import failed. Trying Tk...')
		printW('detail: %s' % e, spaces=True)
		try:	# fail over to the more reliable Tk
			use('TkAgg')
			from tkinter import PhotoImage
		except Exception as e:
			printE('Could not import either Qt or Tk, and the plot module requires at least one of them to run.', sender)
			printE('Please make sure either PyQt5 or Tkinter is installed.', sender, spaces=True)
			printE('detail: %s'% e, sender, spaces=True)
			raise ImportError('Could not import either Qt or Tk, and the plot module requires at least one of them to run')
	import matplotlib.pyplot as plt
	import matplotlib.dates as mdates
	import matplotlib.image as mpimg
	from matplotlib import rcParams
	from matplotlib.ticker import EngFormatter
	rcParams['toolbar'] = 'None'
	plt.ion()
	MPL = True

	# avoiding a matplotlib user warning
	import warnings
	warnings.filterwarnings('ignore', category=UserWarning, module='rsudp')

except Exception as e:
	printE('Could not import matplotlib, plotting will not be available.', sender)
	printE('detail: %s' % e, sender, spaces=True)
	MPL = False

ICON = 'icon.ico'
ICON2 = 'icon.png'

class Plot:
	'''
	.. role:: json(code)
		:language: json

	GUI plotting algorithm, compatible with both Qt5 and TkAgg backends (see :py:func:`matplotlib.use`).
	This module can plot seismogram data from a list of 1-4 Shake channels, and calculate and display a spectrogram beneath each.

	By default the plotted :json:`"duration"` in seconds is :json:`30`.
	The plot will refresh at most once per second, but slower processors may take longer.
	The longer the duration, the more processor power it will take to refresh the plot,
	especially when the spectrogram is enabled.
	To disable the spectrogram, set :json:`"spectrogram"` to :json:`false` in the settings file.
	To put the plot into fullscreen window mode, set :json:`"fullscreen"` to :json:`true`.
	To put the plot into kiosk mode, set :json:`"kiosk"` to :json:`true`.

	:param cha: channels to plot. Defaults to "all" but can be passed a list of channel names as strings.
	:type cha: str or list
	:param int seconds: number of seconds to plot. Defaults to 30.
	:param bool spectrogram: whether to plot the spectrogram. Defaults to True.
	:param bool fullscreen: whether to plot in a fullscreen window. Defaults to False.
	:param bool kiosk: whether to plot in kiosk mode (true fullscreen). Defaults to False.
	:param deconv: whether to deconvolve the signal. Defaults to False.
	:type deconv: str or bool
	:param bool screencap: whether or not to save screenshots of events. Defaults to False.
	:param bool alert: whether to draw the number of events at startup. Defaults to True.
	:param queue.Queue q: queue of data and messages sent by :class:`rsudp.c_consumer.Consumer`
	:raise ImportError: if the module cannot import either of the Matplotlib Qt5 or TkAgg backends
	'''
	def _set_deconv(self, deconv):
		'''
		This function sets the deconvolution units. Allowed values are as follows:

		.. |ms2| replace:: m/s\ :sup:`2`\

		- ``'VEL'`` - velocity (m/s)
		- ``'ACC'`` - acceleration (|ms2|)
		- ``'GRAV'`` - fraction of acceleration due to gravity (g, or 9.81 |ms2|)
		- ``'DISP'`` - displacement (m)
		- ``'CHAN'`` - channel-specific unit calculation, i.e. ``'VEL'`` for geophone channels and ``'ACC'`` for accelerometer channels

		:param str deconv: ``'VEL'``, ``'ACC'``, ``'GRAV'``, ``'DISP'``, or ``'CHAN'``
		'''
		self.deconv = deconv if (deconv in rs.UNITS) else False
		if self.deconv and rs.inv:
			deconv = deconv.upper()
			if self.deconv in rs.UNITS:
				self.units = rs.UNITS[self.deconv][0]
				self.unit = rs.UNITS[self.deconv][1]
			printM('Signal deconvolution set to %s' % (self.deconv), self.sender)
		else:
			self.units = rs.UNITS['CHAN'][0]
			self.unit = rs.UNITS['CHAN'][1]
			self.deconv = False
		printM('Seismogram units are %s' % (self.units), self.sender)


	def __init__(self, q, cha='all',
				 seconds=30, spectrogram=True,
				 fullscreen=False, kiosk=False,
				 deconv=False, screencap=False,
				 alert=True, testing=False):
		"""
		Initialize the plot process.

		"""
		super().__init__()
		self.sender = 'Plot'
		self.alive = True
		self.alarm = False			# don't touch this
		self.alarm_reset = False	# don't touch this

		if MPL == False:
			sys.stdout.flush()
			sys.exit()
		if QT == False:
			printW('Running on %s machine, using Tk instead of Qt' % (platform.machine()), self.sender)

		self.queue = q
		self.master_queue = None	# careful with this, this goes directly to the master consumer. gets set by main thread.

		self.stream = rs.Stream()
		self.raw = rs.Stream()
		self.stn = rs.stn
		self.net = rs.net

		self.chans = []
		helpers.set_channels(self, cha)
		printM('Plotting %s channels: %s' % (len(self.chans), self.chans), self.sender)
		self.totchns = rs.numchns

		self.seconds = seconds
		self.pkts_in_period = rs.tr * rs.numchns * self.seconds	# theoretical number of packets received in self.seconds
		self.spectrogram = spectrogram

		self._set_deconv(deconv)

		self.per_lap = 0.9
		self.fullscreen = fullscreen
		self.kiosk = kiosk
		self.num_chans = len(self.chans)
		self.delay = rs.tr if (self.spectrogram) else 1
		self.delay = 0.5 if (self.chans == ['SHZ']) else self.delay

		self.screencap = screencap
		self.save_timer = 0
		self.save_pct = 0.7
		self.save = []
		self.events = 0
		self.event_text = ' - detected events: 0' if alert else ''
		self.last_event = []
		self.last_event_str = False
		# plot stuff
		self.bgcolor = '#202530' # background
		self.fgcolor = '0.8' # axis and label color
		self.linecolor = '#c28285' # seismogram color

		printM('Starting.', self.sender)

	def deconvolve(self):
		'''
		Send the streams to the central library deconvolve function.
		'''
		helpers.deconvolve(self)

	def getq(self):
		'''
		Get data from the queue and test for whether it has certain strings.
		ALARM and TERM both trigger specific behavior.
		ALARM messages cause the event counter to increment, and if
		:py:data:`screencap==True` then aplot image will be saved when the
		event is :py:data:`self.save_pct` of the way across the plot.
		'''
		d = self.queue.get()
		self.queue.task_done()
		if 'TERM' in str(d):
			plt.close()
			if 'SELF' in str(d):
				printM('Plot has been closed, plot thread will exit.', self.sender)
			self.alive = False
			rs.producer = False

		elif 'ALARM' in str(d):
			self.events += 1		# add event to count
			self.save_timer -= 1	# don't push the save time forward if there are a large number of alarm events
			event = [self.save_timer + int(self.save_pct*self.pkts_in_period),
					 helpers.fsec(helpers.get_msg_time(d))]	# event = [save after count, datetime]
			self.last_event_str = '%s UTC' % (event[1].strftime('%Y-%m-%d %H:%M:%S.%f')[:22])
			printM('Event time: %s' % (self.last_event_str), sender=self.sender)		# show event time in the logs
			if self.screencap:
				printM('Saving png in about %i seconds' % (self.save_pct * (self.seconds)), self.sender)
				self.save.append(event) # append 
			self.fig.suptitle('%s.%s live output - detected events: %s' # title
							% (self.net, self.stn, self.events),
							fontsize=14, color=self.fgcolor, x=0.52)
			self.fig.canvas.set_window_title('(%s) %s.%s - Raspberry Shake Monitor' % (self.events, self.net, self.stn))

		if rs.getCHN(d) in self.chans:
			self.raw = rs.update_stream(
				stream=self.raw, d=d, fill_value='latest')
			return True
		else:
			return False
		
	def set_sps(self):
		'''
		Get samples per second from the main library.
		'''
		self.sps = rs.sps

	# from https://docs.obspy.org/_modules/obspy/imaging/spectrogram.html#_nearest_pow_2:
	def _nearest_pow_2(self, x):
		"""
		Find power of two nearest to x

		>>> _nearest_pow_2(3)
		2.0
		>>> _nearest_pow_2(15)
		16.0

		:type x: float
		:param x: Number
		:rtype: Int
		:return: Nearest power of 2 to x

		Adapted from the `obspy <https://obspy.org>`_ library
		"""
		a = math.pow(2, math.ceil(np.log2(x)))
		b = math.pow(2, math.floor(np.log2(x)))
		if abs(a - x) < abs(b - x):
			return a
		else:
			return b

	def handle_close(self, evt):
		'''
		Handles a plot close event.
		This will trigger a full shutdown of all other processes related to rsudp.
		'''
		self.master_queue.put(helpers.msg_term())

	def handle_resize(self, evt=False):
		'''
		Handles a plot window resize event.
		This will allow the plot to resize dynamically.
		'''
		if evt:
			h = evt.height
		else:
			h = self.fig.get_size_inches()[1]*self.fig.dpi
		plt.tight_layout(pad=0, h_pad=0.1, w_pad=0,
					rect=[0.02, 0.01, 0.98, 0.90 + 0.045*(h/1080)])	# [left, bottom, right, top]

	def _eventsave(self):
		'''
		This function takes the next event in line and pops it out of the list,
		so that it can be saved and others preserved.
		Then, it sets the title to something having to do with the event,
		then calls the save figure function, and finally resets the title.
		'''
		self.save.reverse()
		event = self.save.pop()
		self.save.reverse()

		event_time_str = event[1].strftime('%Y-%m-%d-%H%M%S')				# event time for filename
		title_time_str = event[1].strftime('%Y-%m-%d %H:%M:%S.%f')[:22]		# pretty event time for plot

		# change title (just for a moment)
		self.fig.suptitle('%s.%s detected event - %s UTC' # title
						  % (self.net, self.stn, title_time_str),
						  fontsize=14, color=self.fgcolor, x=0.52)

		# save figure
		self.savefig(event_time=event[1], event_time_str=event_time_str)

		# reset title
		self._set_fig_title()


	def savefig(self, event_time=rs.UTCDateTime.now(),
				event_time_str=rs.UTCDateTime.now().strftime('%Y-%m-%d-%H%M%S')):
		'''
		Saves the figure and puts an IMGPATH message on the master queue.
		This message can be used to upload the image to various services.

		:param obspy.core.utcdatetime.UTCDateTime event_time: Event time as an obspy UTCDateTime object.
		:param str event_time_str: Event time as a string. This is used to set the filename.
		'''
		figname = os.path.join(scap_dir, '%s-%s.png' % (self.stn, event_time_str))
		elapsed = rs.UTCDateTime.now() - event_time
		if int(elapsed) > 0:
			printM('Saving png %i seconds after alarm' % (elapsed), sender=self.sender)
		plt.savefig(figname, facecolor=self.fig.get_facecolor(), edgecolor='none')
		printM('Saved %s' % (figname), sender=self.sender)
		printM('%s thread has saved an image, sending IMGPATH message to queues' % self.sender, sender=self.sender)
		# imgpath requires a UTCDateTime and a string figure path
		self.master_queue.put(helpers.msg_imgpath(event_time, figname))


	def _set_fig_title(self):
		'''
		Sets the figure title back to something that makes sense for the live viewer.
		'''
		self.fig.suptitle('%s.%s live output - detected events: %s' # title
						  % (self.net, self.stn, self.events),
						  fontsize=14, color=self.fgcolor, x=0.52)


	def _init_plot(self):
		'''
		Initialize plot elements and calculate parameters.
		'''
		self.fig = plt.figure(figsize=(11,3*self.num_chans))
		self.fig.canvas.mpl_connect('close_event', self.handle_close)
		self.fig.canvas.mpl_connect('resize_event', self.handle_resize)
		
		if QT:
			self.fig.canvas.window().statusBar().setVisible(False) # remove bottom bar
		self.fig.canvas.set_window_title('%s.%s - Raspberry Shake Monitor' % (self.net, self.stn))
		self.fig.patch.set_facecolor(self.bgcolor)	# background color
		self.fig.suptitle('%s.%s live output%s'	# title
						  % (self.net, self.stn, self.event_text),
						  fontsize=14, color=self.fgcolor,x=0.52)
		self.ax, self.lines = [], []				# list for subplot axes and lines artists
		self.mult = 1					# spectrogram selection multiplier
		if self.spectrogram:
			self.mult = 2				# 2 if user wants a spectrogram else 1
			if self.seconds > 60:
				self.per_lap = 0.9		# if axis is long, spectrogram overlap can be shorter
			else:
				self.per_lap = 0.975	# if axis is short, increase resolution
			# set spectrogram parameters
			self.nfft1 = self._nearest_pow_2(self.sps)
			self.nlap1 = self.nfft1 * self.per_lap


	def _init_axes(self, i):
		'''
		Initialize plot axes.
		'''
		if i == 0:
			# set up first axes (axes added later will share these x axis limits)
			self.ax.append(self.fig.add_subplot(self.num_chans*self.mult,
							1, 1, label=str(1)))
			self.ax[0].set_facecolor(self.bgcolor)
			self.ax[0].tick_params(colors=self.fgcolor, labelcolor=self.fgcolor)
			self.ax[0].xaxis.set_major_formatter(mdates.DateFormatter('%H:%M:%S'))
			self.ax[0].yaxis.set_major_formatter(EngFormatter(unit='%s' % self.unit.lower()))
			if self.spectrogram:
				self.ax.append(self.fig.add_subplot(self.num_chans*self.mult,
								1, 2, label=str(2)))#, sharex=ax[0]))
				self.ax[1].set_facecolor(self.bgcolor)
				self.ax[1].tick_params(colors=self.fgcolor, labelcolor=self.fgcolor)
		else:
			# add axes that share either lines or spectrogram axis limits
			s = i * self.mult	# plot selector
			# add a subplot then set colors
			self.ax.append(self.fig.add_subplot(self.num_chans*self.mult,
							1, s+1, sharex=self.ax[0], label=str(s+1)))
			self.ax[s].set_facecolor(self.bgcolor)
			self.ax[s].tick_params(colors=self.fgcolor, labelcolor=self.fgcolor)
			self.ax[s].xaxis.set_major_formatter(mdates.DateFormatter('%H:%M:%S'))
			self.ax[s].yaxis.set_major_formatter(EngFormatter(unit='%s' % self.unit.lower()))
			if self.spectrogram:
				# add a spectrogram and set colors
				self.ax.append(self.fig.add_subplot(self.num_chans*self.mult,
								1, s+2, sharex=self.ax[1], label=str(s+2)))
				self.ax[s+1].set_facecolor(self.bgcolor)
				self.ax[s+1].tick_params(colors=self.fgcolor, labelcolor=self.fgcolor)


	def _set_icon(self):
		'''
		Set RS plot icons.
		'''
		mgr = plt.get_current_fig_manager()
		ico = pr.resource_filename('rsudp', os.path.join('img', ICON))
		if QT:
			mgr.window.setWindowIcon(QtGui.QIcon(ico))
		else:
			try:
				ico = PhotoImage(file=ico)
				mgr.window.tk.call('wm', 'iconphoto', mgr.window._w, ico)
			except:
				printW('Failed to set PNG icon image, trying .ico instead', sender=self.sender)
				try:
					ico = pr.resource_filename('rsudp', os.path.join('img', ICON2))
					ico = PhotoImage(file=ico)
					mgr.window.tk.call('wm', 'iconphoto', mgr.window._w, ico)
				except:
					printE('Failed to set window icon.')


	def _format_axes(self):
		'''
		Setting up axes and artists.
		'''
		# calculate times
		start = np.datetime64(self.stream[0].stats.endtime
							  )-np.timedelta64(self.seconds, 's')	# numpy time
		end = np.datetime64(self.stream[0].stats.endtime)	# numpy time

		im = mpimg.imread(pr.resource_filename('rsudp', os.path.join('img', 'version1-01-small.png')))
		self.imax = self.fig.add_axes([0.015, 0.944, 0.2, 0.056], anchor='NW') # [left, bottom, right, top]
		self.imax.imshow(im, aspect='equal', interpolation='sinc')
		self.imax.axis('off')
		# set up axes and artists
		for i in range(self.num_chans): # create lines objects and modify axes
			if len(self.stream[i].data) < int(self.sps*(1/self.per_lap)):
				comp = 0				# spectrogram offset compensation factor
			else:
				comp = (1/self.per_lap)**2	# spectrogram offset compensation factor
			r = np.arange(start, end, np.timedelta64(int(1000/self.sps), 'ms'))[-len(
						  self.stream[i].data[int(-self.sps*(self.seconds-(comp/2))):-int(self.sps*(comp/2))]):]
			mean = int(round(np.mean(self.stream[i].data)))
			# add artist to lines list
			self.lines.append(self.ax[i*self.mult].plot(r,
							  np.nan*(np.zeros(len(r))),
							  label=self.stream[i].stats.channel, color=self.linecolor,
							  lw=0.45)[0])
			# set axis limits
			self.ax[i*self.mult].set_xlim(left=start.astype(datetime),
										  right=end.astype(datetime))
			self.ax[i*self.mult].set_ylim(bottom=np.min(self.stream[i].data-mean)
										  -np.ptp(self.stream[i].data-mean)*0.1,
										  top=np.max(self.stream[i].data-mean)
										  +np.ptp(self.stream[i].data-mean)*0.1)
			# we can set line plot labels here, but not imshow labels
			ylabel = self.stream[i].stats.units.strip().capitalize() if (' ' in self.stream[i].stats.units) else self.stream[i].stats.units
			self.ax[i*self.mult].set_ylabel(ylabel, color=self.fgcolor)
			self.ax[i*self.mult].legend(loc='upper left')	# legend and location
			if self.spectrogram:		# if the user wants a spectrogram, plot it
				# add spectrogram to axes list
				sg = self.ax[1].specgram(self.stream[i].data, NFFT=8, pad_to=8,
										 Fs=self.sps, noverlap=7, cmap='inferno',
										 xextent=(self.seconds-0.5, self.seconds))[0]
				self.ax[1].set_xlim(0,self.seconds)
				self.ax[i*self.mult+1].set_ylim(0,int(self.sps/2))
				self.ax[i*self.mult+1].imshow(np.flipud(sg**(1/float(10))), cmap='inferno',
						extent=(self.seconds-(1/(self.sps/float(len(self.stream[i].data)))),
								self.seconds,0,self.sps/2), aspect='auto')


	def _setup_fig_manager(self):
		'''
		Setting up figure manager and 
		'''
		# update canvas and draw
		figManager = plt.get_current_fig_manager()
		if self.kiosk:
			figManager.full_screen_toggle()
		else:
			if self.fullscreen:	# set fullscreen
				if QT:	# maximizing in Qt
					figManager.window.showMaximized()
				else:	# maximizing in Tk
					figManager.resize(*figManager.window.maxsize())


	def setup_plot(self):
		"""
		Sets up the plot. Quite a lot of stuff happens in this function.
		Matplotlib backends are not threadsafe, so things are a little weird.
		See code comments for details.
		"""
		# instantiate a figure and set basic params
		self._init_plot()

		for i in range(self.num_chans):
			self._init_axes(i)

		for axis in self.ax:
			# set the rest of plot colors
			plt.setp(axis.spines.values(), color=self.fgcolor)
			plt.setp([axis.get_xticklines(), axis.get_yticklines()], color=self.fgcolor)

		# rs logos
		self._set_icon()

		# draw axes
		self._format_axes()

		self.handle_resize()

		# setup figure manager
		self._setup_fig_manager()

		# draw plot, loop, and resize the plot
		plt.draw()									# draw the canvas
		self.fig.canvas.start_event_loop(0.005)		# wait for canvas to update
		self.handle_resize()


	def _set_ch_specific_label(self, i):
		'''
		Set the formatter units if the deconvolution is channel-specific.
		'''
		if self.deconv:
			if (self.deconv in 'CHAN'):
				ch = self.stream[i].stats.channel
				if ('HZ' in ch) or ('HN' in ch) or ('HE' in ch):
					unit = rs.UNITS['VEL'][1]
				elif ('EN' in ch):
					unit = rs.UNITS['ACC'][1]
				else:
					unit = rs.UNITS['CHAN'][1]
				self.ax[i*self.mult].yaxis.set_major_formatter(EngFormatter(unit='%s' % unit.lower()))


	def _draw_lines(self, i, start, end, mean):
		'''
		Updates the line data in the plot.

		:param int i: the trace number
		:param numpy.datetime64 start: start time of the trace
		:param numpy.datetime64 end: end time of the trace
		:param float mean: the mean of data in the trace
		'''
		comp = 1/self.per_lap	# spectrogram offset compensation factor
		r = np.arange(start, end, np.timedelta64(int(1000/self.sps), 'ms'))[-len(
					self.stream[i].data[int(-self.sps*(self.seconds-(comp/2))):-int(self.sps*(comp/2))]):]
		self.lines[i].set_ydata(self.stream[i].data[int(-self.sps*(self.seconds-(comp/2))):-int(self.sps*(comp/2))]-mean)
		self.lines[i].set_xdata(r)	# (1/self.per_lap)/2
		self.ax[i*self.mult].set_xlim(left=start.astype(datetime)+timedelta(seconds=comp*1.5),
										right=end.astype(datetime))
		self.ax[i*self.mult].set_ylim(bottom=np.min(self.stream[i].data-mean)
										-np.ptp(self.stream[i].data-mean)*0.1,
										top=np.max(self.stream[i].data-mean)
										+np.ptp(self.stream[i].data-mean)*0.1)


	def _update_specgram(self, i, mean):
		'''
		Updates the spectrogram and its labels.

		:param int i: the trace number
		:param float mean: the mean of data in the trace
		'''
		self.nfft1 = self._nearest_pow_2(self.sps)	# FFTs run much faster if the number of transforms is a power of 2
		self.nlap1 = self.nfft1 * self.per_lap
		if len(self.stream[i].data) < self.nfft1:	# when the number of data points is low, we just need to kind of fake it for a few fractions of a second
			self.nfft1 = 8
			self.nlap1 = 6
		sg = self.ax[i*self.mult+1].specgram(self.stream[i].data-mean,
					NFFT=self.nfft1, pad_to=int(self.nfft1*4), # previously self.sps*4),
					Fs=self.sps, noverlap=self.nlap1)[0]	# meat & potatoes
		self.ax[i*self.mult+1].clear()	# incredibly important, otherwise continues to draw over old images (gets exponentially slower)
		# cloogy way to shift the spectrogram to line up with the seismogram
		self.ax[i*self.mult+1].set_xlim(0.25,self.seconds-0.25)
		self.ax[i*self.mult+1].set_ylim(0,int(self.sps/2))
		# imshow to update the spectrogram
		self.ax[i*self.mult+1].imshow(np.flipud(sg**(1/float(10))), cmap='inferno',
				extent=(self.seconds-(1/(self.sps/float(len(self.stream[i].data)))),
						self.seconds,0,self.sps/2), aspect='auto')
		# some things that unfortunately can't be in the setup function:
		self.ax[i*self.mult+1].tick_params(axis='x', which='both',
				bottom=False, top=False, labelbottom=False)
		self.ax[i*self.mult+1].set_ylabel('Frequency (Hz)', color=self.fgcolor)
		self.ax[i*self.mult+1].set_xlabel('Time (UTC)', color=self.fgcolor)


	def update_plot(self):
		'''
		Redraw the plot with new data.
		Called on every nth loop after the plot is set up, where n is
		the number of channels times the data packet arrival rate in Hz.
		This has the effect of making the plot update once per second.
		'''
		obstart = self.stream[0].stats.endtime - timedelta(seconds=self.seconds)	# obspy time
		start = np.datetime64(self.stream[0].stats.endtime
							  )-np.timedelta64(self.seconds, 's')	# numpy time
		end = np.datetime64(self.stream[0].stats.endtime)	# numpy time
		self.raw = self.raw.slice(starttime=obstart)	# slice the stream to the specified length (seconds variable)
		self.stream = self.stream.slice(starttime=obstart)	# slice the stream to the specified length (seconds variable)
		i = 0
		for i in range(self.num_chans):	# for each channel, update the plots
			mean = int(round(np.mean(self.stream[i].data)))
			self._draw_lines(i, start, end, mean)
			self._set_ch_specific_label(i)
			if self.spectrogram:
				self._update_specgram(i, mean)
			else:
				# also can't be in the setup function
				self.ax[i*self.mult].set_xlabel('Time (UTC)', color=self.fgcolor)


	def figloop(self):
		"""
		Let some time elapse in order for the plot canvas to draw properly.
		Must be separate from :py:func:`update_plot()` to avoid a broadcast error early in plotting.
		"""
		self.fig.canvas.start_event_loop(0.005)


	def mainloop(self, i, u):
		'''
		The main loop in the :py:func:`rsudp.c_plot.Plot.run`.

		:param int i: number of plot events without clearing the linecache
		:param int u: queue blocking counter
		:return: number of plot events without clearing the linecache and queue blocking counter
		:rtype: int, int
		'''
		if i > 10:
			linecache.clearcache()
			i = 0
		else:
			i += 1
		self.stream = rs.copy(self.stream)	# essential, otherwise the stream has a memory leak
		self.raw = rs.copy(self.raw)		# and could eventually crash the machine
		self.deconvolve()
		self.update_plot()
		if u >= 0:				# avoiding a matplotlib broadcast error
			self.figloop()

		if self.save:
			# save the plot
			if (self.save_timer > self.save[0][0]):
				self._eventsave()
		u = 0
		time.sleep(0.005)		# wait a ms to see if another packet will arrive
		sys.stdout.flush()
		return i, u

	def qu(self, u):
		'''
		Get a queue object and increment the queue counter.
		This is a way to figure out how many channels have arrived in the queue.

		:param int u: queue blocking counter
		:return: queue blocking counter
		:rtype: int
		'''
		u += 1 if self.getq() else 0
		return u


	def run(self):
		"""
		The heart of the plotting routine.

		Begins by updating the queue to populate a :py:class:`obspy.core.stream.Stream` object, then setting up the main plot.
		The first time through the main loop, the plot is not drawn. After that, the plot is drawn every time all channels are updated.
		Any plots containing a spectrogram and more than 1 channel are drawn at most every second (1000 ms).
		All other plots are drawn at most every quarter second (250 ms).
		"""
		self.getq() # block until data is flowing from the consumer
		for i in range((self.totchns)*2): # fill up a stream object
			self.getq()
		self.set_sps()
		self.deconvolve()
		self.setup_plot()

		n = 0	# number of iterations without plotting
		i = 0	# number of plot events without clearing the linecache
		u = -1	# number of blocked queue calls (must be -1 at startup)
		while True: # main loop
			while True: # sub loop
				if self.alive == False:	# break if the user has closed the plot
					break
				n += 1
				self.save_timer += 1
				if self.queue.qsize() > 0:
					self.getq()
					time.sleep(0.009)		# wait a ms to see if another packet will arrive
				else:
					u = self.qu(u)
					if n > (self.delay * rs.numchns):
						n = 0
						break
			if self.alive == False:	# break if the user has closed the plot
				printM('Exiting.', self.sender)
				break
			i, u = self.mainloop(i, u)
			if self.testing:
				TEST['c_plot'][1] = True
		return

raspishake / rsudp

Push — coverage ( 0196fb...b39e1a )

build.rsudp.c_plot.Plot.run() C

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like
