New orientation listener.

The objective in this listener is to be more careful about the
signal processing to prevent spurious orientation changes
and to make all of the tweakable factors physically meaningful.

The calibration is defined in terms of time constants and
does not assume a particular discrete sampling rate.  This is
useful because it allows us to change the accelerometer sampling
interval if desired without having to change the calibration.
Moreover, the accelerometer sampling interval can vary +/- 20ms
from one sample to the next even in normal circumstances.

Proposed orientation changes are weighted by confidence factors
that vary exponentially in relation to how close the device
is to the ideal orientation change posture (screen is vertical,
angle is exactly at the midpoint of the orientation quadrant,
and no external acceleration beside gravity).  When not in an ideal
posture, the device takes proportionally longer to settle into a
new orientation state.

Added a little tool to plot the log output of the
WindowOrientationListener.  Check the README for more information
about how to use it.

Change-Id: I787f02d03582ff26367df65eda8d9ce85c5cb343

2 files changed, 538 insertions, 0 deletions

diff --git a/tools/orientationplot/README.txt b/tools/orientationplot/README.txt
new file mode 100644
index 0000000..0143510
--- /dev/null
+++ b/tools/orientationplot/README.txt
@@ -0,0 +1,87 @@
+This directory contains a simple python script for visualizing
+the behavior of the WindowOrientationListener.
+
+
+PREREQUISITES
+-------------
+
+1. Python 2.6
+2. numpy
+3. matplotlib
+
+
+USAGE
+-----
+
+The tool works by scaping the debug log output from WindowOrientationListener
+for interesting data and then plotting it.
+
+1. Enable the Window Orientation Listener debugging data log using the
+   Development Settings in the Dev Tools application (Development.apk).
+
+2. Plug in the device.  Ensure that it is the only device plugged in
+   since this script is of very little brain and will get confused otherwise.
+
+3. Run "orientationplot.py".
+
+4. When finished, remember to disable the debug log output since it is quite verbose!
+
+
+WHAT IT ALL MEANS
+-----------------
+
+The tool displays several time series graphs that plot the output of the
+WindowOrientationListener.  Here you can see the raw accelerometer data,
+filtered accelerometer data, measured tilt and orientation angle, confidence
+intervals for the proposed orientation and accelerometer latency.
+
+Things to look for:
+
+1. Ensure the filtering is not too aggressive.  If the filter cut-off frequency is
+   less than about 1Hz, then the filtered accelorometer data becomes too smooth
+   and the latency for orientation detection goes up.  One way to observe this
+   is by holding the device vertically in one orientation then sharply turning
+   it 90 degrees to a different orientation.  Compared the rapid changes in the
+   raw accelerometer data with the smoothed out filtered data.  If the filtering
+   is too aggressive, the filter response may lag by hundreds of milliseconds.
+
+2. Ensure that there is an appropriate gap between adjacent orientation angles
+   for hysteresis.  Try holding the device in one orientation and slowly turning
+   it 90 degrees.  Note that the confidence intervals will all drop to 0 at some
+   point in between the two orientations; that is the gap.  The gap should be
+   observed between all adjacent pairs of orientations when turning the device
+   in either direction.
+
+   Next try holding the device in one orientation and rapidly turning it end
+   over end to a midpoint about 45 degrees between two opposing orientations.
+   There should be no gap observed initially.  The algorithm should pick one
+   of the orientations and settle into it (since it is obviously quite
+   different from the original orientation of the device).  However, once it
+   settles, the confidence values should start trending to 0 again because
+   the measured orientation angle is now within the gap between the new
+   orientation and the adjacent orientation.
+
+   In other words, the hysteresis gap applies only when the measured orientation
+   angle (say, 45 degrees) is between the current orientation's ideal angle
+   (say, 0 degrees) and an adjacent orientation's ideal angle (say, 90 degrees).
+
+3. Accelerometer jitter.  The accelerometer latency graph displays the interval
+   between sensor events as reported by the SensorEvent.timestamp field.  It
+   should be a fairly constant 60ms.  If the latency jumps around wildly or
+   greatly exceeds 60ms then there is a problem with the accelerometer or the
+   sensor manager.
+
+4. The orientation angle is not measured when the tilt is too close to 90 or -90
+   degrees (refer to MAX_TILT constant).  Consequently, you should expect there
+   to be no data.  Likewise, all dependent calculations are suppressed in this case
+   so there will be no orientation proposal either.
+
+5. Each orientation has its own bound on allowable tilt angles.  It's a good idea to
+   verify that these limits are being enforced by gradually varying the tilt of
+   the device until it is inside/outside the limit for each orientation.
+
+6. Orientation changes should be significantly harder when the device is held
+   overhead.  People reading on tablets in bed often have their head turned
+   a little to the side, or they hold the device loosely so its orientation
+   can be a bit unusual.  The tilt is a good indicator of whether the device is
+   overhead.
diff --git a/tools/orientationplot/orientationplot.py b/tools/orientationplot/orientationplot.py
new file mode 100755
index 0000000..07449d4
--- /dev/null
+++ b/tools/orientationplot/orientationplot.py
@@ -0,0 +1,451 @@
+#!/usr/bin/env python2.6
+#
+# Copyright (C) 2011 The Android Open Source Project
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+#
+# Plots debug log output from WindowOrientationListener.
+# See README.txt for details.
+#
+
+import numpy as np
+import matplotlib.pyplot as plot
+import subprocess
+import re
+import fcntl
+import os
+import errno
+import bisect
+from datetime import datetime, timedelta
+
+# Parameters.
+timespan = 15 # seconds total span shown
+scrolljump = 5 # seconds jump when scrolling
+timeticks = 1 # seconds between each time tick
+
+# Non-blocking stream wrapper.
+class NonBlockingStream:
+  def __init__(self, stream):
+    fcntl.fcntl(stream, fcntl.F_SETFL, os.O_NONBLOCK)
+    self.stream = stream
+    self.buffer = ''
+    self.pos = 0
+
+  def readline(self):
+    while True:
+      index = self.buffer.find('\n', self.pos)
+      if index != -1:
+        result = self.buffer[self.pos:index]
+        self.pos = index + 1
+        return result
+
+      self.buffer = self.buffer[self.pos:]
+      self.pos = 0
+      try:
+        chunk = os.read(self.stream.fileno(), 4096)
+      except OSError, e:
+        if e.errno == errno.EAGAIN:
+          return None
+        raise e
+      if len(chunk) == 0:
+        if len(self.buffer) == 0:
+          raise(EOFError)
+        else:
+          result = self.buffer
+          self.buffer = ''
+          self.pos = 0
+          return result
+      self.buffer += chunk
+
+# Plotter
+class Plotter:
+  def __init__(self, adbout):
+    self.adbout = adbout
+
+    self.fig = plot.figure(1)
+    self.fig.suptitle('Window Orientation Listener', fontsize=12)
+    self.fig.set_dpi(96)
+    self.fig.set_size_inches(16, 12, forward=True)
+
+    self.raw_acceleration_x = self._make_timeseries()
+    self.raw_acceleration_y = self._make_timeseries()
+    self.raw_acceleration_z = self._make_timeseries()
+    self.raw_acceleration_axes = self._add_timeseries_axes(
+        1, 'Raw Acceleration', 'm/s^2', [-20, 20],
+        yticks=range(-15, 16, 5))
+    self.raw_acceleration_line_x = self._add_timeseries_line(
+        self.raw_acceleration_axes, 'x', 'red')
+    self.raw_acceleration_line_y = self._add_timeseries_line(
+        self.raw_acceleration_axes, 'y', 'green')
+    self.raw_acceleration_line_z = self._add_timeseries_line(
+        self.raw_acceleration_axes, 'z', 'blue')
+    self._add_timeseries_legend(self.raw_acceleration_axes)
+
+    shared_axis = self.raw_acceleration_axes
+
+    self.filtered_acceleration_x = self._make_timeseries()
+    self.filtered_acceleration_y = self._make_timeseries()
+    self.filtered_acceleration_z = self._make_timeseries()
+    self.magnitude = self._make_timeseries()
+    self.filtered_acceleration_axes = self._add_timeseries_axes(
+        2, 'Filtered Acceleration', 'm/s^2', [-20, 20],
+        sharex=shared_axis,
+        yticks=range(-15, 16, 5))
+    self.filtered_acceleration_line_x = self._add_timeseries_line(
+        self.filtered_acceleration_axes, 'x', 'red')
+    self.filtered_acceleration_line_y = self._add_timeseries_line(
+        self.filtered_acceleration_axes, 'y', 'green')
+    self.filtered_acceleration_line_z = self._add_timeseries_line(
+        self.filtered_acceleration_axes, 'z', 'blue')
+    self.magnitude_line = self._add_timeseries_line(
+        self.filtered_acceleration_axes, 'magnitude', 'orange', linewidth=2)
+    self._add_timeseries_legend(self.filtered_acceleration_axes)
+
+    self.tilt_angle = self._make_timeseries()
+    self.tilt_angle_axes = self._add_timeseries_axes(
+        3, 'Tilt Angle', 'degrees', [-105, 105],
+        sharex=shared_axis,
+        yticks=range(-90, 91, 30))
+    self.tilt_angle_line = self._add_timeseries_line(
+        self.tilt_angle_axes, 'tilt', 'black')
+    self._add_timeseries_legend(self.tilt_angle_axes)
+
+    self.orientation_angle = self._make_timeseries()
+    self.orientation_angle_axes = self._add_timeseries_axes(
+        4, 'Orientation Angle', 'degrees', [-25, 375],
+        sharex=shared_axis,
+        yticks=range(0, 361, 45))
+    self.orientation_angle_line = self._add_timeseries_line(
+        self.orientation_angle_axes, 'orientation', 'black')
+    self._add_timeseries_legend(self.orientation_angle_axes)
+
+    self.actual_orientation = self._make_timeseries()
+    self.proposed_orientation = self._make_timeseries()
+    self.orientation_axes = self._add_timeseries_axes(
+        5, 'Actual / Proposed Orientation and Confidence', 'rotation', [-1, 4],
+        sharex=shared_axis,
+        yticks=range(0, 4))
+    self.actual_orientation_line = self._add_timeseries_line(
+        self.orientation_axes, 'actual', 'black', linewidth=2)
+    self.proposed_orientation_line = self._add_timeseries_line(
+        self.orientation_axes, 'proposed', 'purple', linewidth=3)
+    self._add_timeseries_legend(self.orientation_axes)
+
+    self.confidence = [[self._make_timeseries(), self._make_timeseries()] for i in range(0, 4)]
+    self.confidence_polys = []
+
+    self.combined_confidence = self._make_timeseries()
+    self.orientation_confidence = self._make_timeseries()
+    self.tilt_confidence = self._make_timeseries()
+    self.magnitude_confidence = self._make_timeseries()
+    self.confidence_axes = self._add_timeseries_axes(
+        6, 'Proposed Orientation Confidence Factors', 'confidence', [-0.1, 1.1],
+        sharex=shared_axis,
+        yticks=[0.0, 0.2, 0.4, 0.6, 0.8, 1.0])
+    self.combined_confidence_line = self._add_timeseries_line(
+        self.confidence_axes, 'combined', 'purple', linewidth=2)
+    self.orientation_confidence_line = self._add_timeseries_line(
+        self.confidence_axes, 'orientation', 'black')
+    self.tilt_confidence_line = self._add_timeseries_line(
+        self.confidence_axes, 'tilt', 'brown')
+    self.magnitude_confidence_line = self._add_timeseries_line(
+        self.confidence_axes, 'magnitude', 'orange')
+    self._add_timeseries_legend(self.confidence_axes)
+
+    self.sample_latency = self._make_timeseries()
+    self.sample_latency_axes = self._add_timeseries_axes(
+        7, 'Accelerometer Sampling Latency', 'ms', [-10, 500],
+        sharex=shared_axis,
+        yticks=range(0, 500, 100))
+    self.sample_latency_line = self._add_timeseries_line(
+        self.sample_latency_axes, 'latency', 'black')
+    self._add_timeseries_legend(self.sample_latency_axes)
+
+    self.timer = self.fig.canvas.new_timer(interval=100)
+    self.timer.add_callback(lambda: self.update())
+    self.timer.start()
+
+    self.timebase = None
+    self._reset_parse_state()
+
+  # Initialize a time series.
+  def _make_timeseries(self):
+    return [[], []]
+
+  # Add a subplot to the figure for a time series.
+  def _add_timeseries_axes(self, index, title, ylabel, ylim, yticks, sharex=None):
+    num_graphs = 7
+    height = 0.9 / num_graphs
+    top = 0.95 - height * index
+    axes = self.fig.add_axes([0.1, top, 0.8, height],
+        xscale='linear',
+        xlim=[0, timespan],
+        ylabel=ylabel,
+        yscale='linear',
+        ylim=ylim,
+        sharex=sharex)
+    axes.text(0.02, 0.02, title, transform=axes.transAxes, fontsize=10, fontweight='bold')
+    axes.set_xlabel('time (s)', fontsize=10, fontweight='bold')
+    axes.set_ylabel(ylabel, fontsize=10, fontweight='bold')
+    axes.set_xticks(range(0, timespan + 1, timeticks))
+    axes.set_yticks(yticks)
+    axes.grid(True)
+
+    for label in axes.get_xticklabels():
+      label.set_fontsize(9)
+    for label in axes.get_yticklabels():
+      label.set_fontsize(9)
+
+    return axes
+
+  # Add a line to the axes for a time series.
+  def _add_timeseries_line(self, axes, label, color, linewidth=1):
+    return axes.plot([], label=label, color=color, linewidth=linewidth)[0]
+
+  # Add a legend to a time series.
+  def _add_timeseries_legend(self, axes):
+    axes.legend(
+        loc='upper left',
+        bbox_to_anchor=(1.01, 1),
+        borderpad=0.1,
+        borderaxespad=0.1,
+        prop={'size': 10})
+
+  # Resets the parse state.
+  def _reset_parse_state(self):
+    self.parse_raw_acceleration_x = None
+    self.parse_raw_acceleration_y = None
+    self.parse_raw_acceleration_z = None
+    self.parse_filtered_acceleration_x = None
+    self.parse_filtered_acceleration_y = None
+    self.parse_filtered_acceleration_z = None
+    self.parse_magnitude = None
+    self.parse_tilt_angle = None
+    self.parse_orientation_angle = None
+    self.parse_proposed_orientation = None
+    self.parse_combined_confidence = None
+    self.parse_orientation_confidence = None
+    self.parse_tilt_confidence = None
+    self.parse_magnitude_confidence = None
+    self.parse_actual_orientation = None
+    self.parse_confidence = None
+    self.parse_sample_latency = None
+
+  # Update samples.
+  def update(self):
+    timeindex = 0
+    while True:
+      try:
+        line = self.adbout.readline()
+      except EOFError:
+        plot.close()
+        return
+      if line is None:
+        break
+      print line
+
+      try:
+        timestamp = self._parse_timestamp(line)
+      except ValueError, e:
+        continue
+      if self.timebase is None:
+        self.timebase = timestamp
+      delta = timestamp - self.timebase
+      timeindex = delta.seconds + delta.microseconds * 0.000001
+
+      if line.find('Raw acceleration vector:') != -1:
+        self.parse_raw_acceleration_x = self._get_following_number(line, 'x=')
+        self.parse_raw_acceleration_y = self._get_following_number(line, 'y=')
+        self.parse_raw_acceleration_z = self._get_following_number(line, 'z=')
+
+      if line.find('Filtered acceleration vector:') != -1:
+        self.parse_filtered_acceleration_x = self._get_following_number(line, 'x=')
+        self.parse_filtered_acceleration_y = self._get_following_number(line, 'y=')
+        self.parse_filtered_acceleration_z = self._get_following_number(line, 'z=')
+
+      if line.find('magnitude=') != -1:
+        self.parse_magnitude = self._get_following_number(line, 'magnitude=')
+
+      if line.find('tiltAngle=') != -1:
+        self.parse_tilt_angle = self._get_following_number(line, 'tiltAngle=')
+
+      if line.find('orientationAngle=') != -1:
+        self.parse_orientation_angle = self._get_following_number(line, 'orientationAngle=')
+
+      if line.find('Proposal:') != -1:
+        self.parse_proposed_orientation = self._get_following_number(line, 'proposedOrientation=')
+        self.parse_combined_confidence = self._get_following_number(line, 'combinedConfidence=')
+        self.parse_orientation_confidence = self._get_following_number(line, 'orientationConfidence=')
+        self.parse_tilt_confidence = self._get_following_number(line, 'tiltConfidence=')
+        self.parse_magnitude_confidence = self._get_following_number(line, 'magnitudeConfidence=')
+
+      if line.find('Result:') != -1:
+        self.parse_actual_orientation = self._get_following_number(line, 'rotation=')
+        self.parse_confidence = self._get_following_array_of_numbers(line, 'confidence=')
+        self.parse_sample_latency = self._get_following_number(line, 'timeDeltaMS=')
+
+        for i in range(0, 4):
+          if self.parse_confidence is not None:
+            self._append(self.confidence[i][0], timeindex, i)
+            self._append(self.confidence[i][1], timeindex, i + self.parse_confidence[i])
+          else:
+            self._append(self.confidence[i][0], timeindex, None)
+            self._append(self.confidence[i][1], timeindex, None)
+
+        self._append(self.raw_acceleration_x, timeindex, self.parse_raw_acceleration_x)
+        self._append(self.raw_acceleration_y, timeindex, self.parse_raw_acceleration_y)
+        self._append(self.raw_acceleration_z, timeindex, self.parse_raw_acceleration_z)
+        self._append(self.filtered_acceleration_x, timeindex, self.parse_filtered_acceleration_x)
+        self._append(self.filtered_acceleration_y, timeindex, self.parse_filtered_acceleration_y)
+        self._append(self.filtered_acceleration_z, timeindex, self.parse_filtered_acceleration_z)
+        self._append(self.magnitude, timeindex, self.parse_magnitude)
+        self._append(self.tilt_angle, timeindex, self.parse_tilt_angle)
+        self._append(self.orientation_angle, timeindex, self.parse_orientation_angle)
+        self._append(self.actual_orientation, timeindex, self.parse_actual_orientation)
+        self._append(self.proposed_orientation, timeindex, self.parse_proposed_orientation)
+        self._append(self.combined_confidence, timeindex, self.parse_combined_confidence)
+        self._append(self.orientation_confidence, timeindex, self.parse_orientation_confidence)
+        self._append(self.tilt_confidence, timeindex, self.parse_tilt_confidence)
+        self._append(self.magnitude_confidence, timeindex, self.parse_magnitude_confidence)
+        self._append(self.sample_latency, timeindex, self.parse_sample_latency)
+        self._reset_parse_state()
+
+    # Scroll the plots.
+    if timeindex > timespan:
+      bottom = int(timeindex) - timespan + scrolljump
+      self.timebase += timedelta(seconds=bottom)
+      self._scroll(self.raw_acceleration_x, bottom)
+      self._scroll(self.raw_acceleration_y, bottom)
+      self._scroll(self.raw_acceleration_z, bottom)
+      self._scroll(self.filtered_acceleration_x, bottom)
+      self._scroll(self.filtered_acceleration_y, bottom)
+      self._scroll(self.filtered_acceleration_z, bottom)
+      self._scroll(self.magnitude, bottom)
+      self._scroll(self.tilt_angle, bottom)
+      self._scroll(self.orientation_angle, bottom)
+      self._scroll(self.actual_orientation, bottom)
+      self._scroll(self.proposed_orientation, bottom)
+      self._scroll(self.combined_confidence, bottom)
+      self._scroll(self.orientation_confidence, bottom)
+      self._scroll(self.tilt_confidence, bottom)
+      self._scroll(self.magnitude_confidence, bottom)
+      self._scroll(self.sample_latency, bottom)
+      for i in range(0, 4):
+        self._scroll(self.confidence[i][0], bottom)
+        self._scroll(self.confidence[i][1], bottom)
+
+    # Redraw the plots.
+    self.raw_acceleration_line_x.set_data(self.raw_acceleration_x)
+    self.raw_acceleration_line_y.set_data(self.raw_acceleration_y)
+    self.raw_acceleration_line_z.set_data(self.raw_acceleration_z)
+    self.filtered_acceleration_line_x.set_data(self.filtered_acceleration_x)
+    self.filtered_acceleration_line_y.set_data(self.filtered_acceleration_y)
+    self.filtered_acceleration_line_z.set_data(self.filtered_acceleration_z)
+    self.magnitude_line.set_data(self.magnitude)
+    self.tilt_angle_line.set_data(self.tilt_angle)
+    self.orientation_angle_line.set_data(self.orientation_angle)
+    self.actual_orientation_line.set_data(self.actual_orientation)
+    self.proposed_orientation_line.set_data(self.proposed_orientation)
+    self.combined_confidence_line.set_data(self.combined_confidence)
+    self.orientation_confidence_line.set_data(self.orientation_confidence)
+    self.tilt_confidence_line.set_data(self.tilt_confidence)
+    self.magnitude_confidence_line.set_data(self.magnitude_confidence)
+    self.sample_latency_line.set_data(self.sample_latency)
+
+    for poly in self.confidence_polys:
+      poly.remove()
+    self.confidence_polys = []
+    for i in range(0, 4):
+      self.confidence_polys.append(self.orientation_axes.fill_between(self.confidence[i][0][0],
+        self.confidence[i][0][1], self.confidence[i][1][1],
+        facecolor='goldenrod', edgecolor='goldenrod'))
+
+    self.fig.canvas.draw_idle()
+
+  # Scroll a time series.
+  def _scroll(self, timeseries, bottom):
+    bottom_index = bisect.bisect_left(timeseries[0], bottom)
+    del timeseries[0][:bottom_index]
+    del timeseries[1][:bottom_index]
+    for i, timeindex in enumerate(timeseries[0]):
+      timeseries[0][i] = timeindex - bottom
+
+  # Extract a word following the specified prefix.
+  def _get_following_word(self, line, prefix):
+    prefix_index = line.find(prefix)
+    if prefix_index == -1:
+      return None
+    start_index = prefix_index + len(prefix)
+    delim_index = line.find(',', start_index)
+    if delim_index == -1:
+      return line[start_index:]
+    else:
+      return line[start_index:delim_index]
+
+  # Extract a number following the specified prefix.
+  def _get_following_number(self, line, prefix):
+    word = self._get_following_word(line, prefix)
+    if word is None:
+      return None
+    return float(word)
+
+  # Extract an array of numbers following the specified prefix.
+  def _get_following_array_of_numbers(self, line, prefix):
+    prefix_index = line.find(prefix + '[')
+    if prefix_index == -1:
+      return None
+    start_index = prefix_index + len(prefix) + 1
+    delim_index = line.find(']', start_index)
+    if delim_index == -1:
+      return None
+
+    result = []
+    while start_index < delim_index:
+      comma_index = line.find(', ', start_index, delim_index)
+      if comma_index == -1:
+        result.append(float(line[start_index:delim_index]))
+        break;
+      result.append(float(line[start_index:comma_index]))
+      start_index = comma_index + 2
+    return result
+
+  # Add a value to a time series.
+  def _append(self, timeseries, timeindex, number):
+    timeseries[0].append(timeindex)
+    timeseries[1].append(number)
+
+  # Parse the logcat timestamp.
+  # Timestamp has the form '01-21 20:42:42.930'
+  def _parse_timestamp(self, line):
+    return datetime.strptime(line[0:18], '%m-%d %H:%M:%S.%f')
+
+# Notice
+print "Window Orientation Listener plotting tool"
+print "-----------------------------------------\n"
+print "Please turn on the Window Orientation Listener logging in Development Settings."
+
+# Start adb.
+print "Starting adb logcat.\n"
+
+adb = subprocess.Popen(['adb', 'logcat', '-s', '-v', 'time', 'WindowOrientationListener:V'],
+    stdout=subprocess.PIPE)
+adbout = NonBlockingStream(adb.stdout)
+
+# Prepare plotter.
+plotter = Plotter(adbout)
+plotter.update()
+
+# Main loop.
+plot.show()