def dead_reckon(ax, ay, az, t):
"""
Attempt dead reckoning (estimating position) from accelerometer data.
The accelerometer data are too noisy!
Parameters
----------
ax : list or numpy array of floats
accelerometer x-axis data
ay : list or numpy array of floats
accelerometer y-axis data
az : list or numpy array of floats
accelerometer z-axis data
t : list or numpy array of floats
accelerometer time points
Returns
-------
x : list or numpy array of floats
estimated position along x-axis
y : list or numpy array of floats
estimated position along y-axis
z : list or numpy array of floats
estimated position along z-axis
distance : float
estimated change in position
Examples
--------
>>> from mhealthx.extractors.dead_reckon import dead_reckon
>>> from mhealthx.xio import read_accel_json
>>> #input_file = '/Users/arno/DriveWork/mhealthx/mpower_sample_data/accel_walking_outbound.json.items-6dc4a144-55c3-4e6d-982c-19c7a701ca243282023468470322798.tmp'
>>> input_file = '/Users/arno/DriveWork/mhealthx/mpower_sample_data/deviceMotion_walking_outbound.json.items-5981e0a8-6481-41c8-b589-fa207bfd2ab38771455825726024828.tmp'
>>> start = 150
>>> device_motion = True
>>> t, axyz, gxyz, uxyz, rxyz, sample_rate, duration = read_accel_json(input_file, start, device_motion)
>>> ax, ay, az = axyz
>>> x, y, z, distance = dead_reckon(ax, ay, az, t)
"""
import numpy as np
from mhealthx.extractors.dead_reckon import velocity_from_acceleration,\
position_from_velocity
#-------------------------------------------------------------------------
# De-mean accelerometer readings:
#-------------------------------------------------------------------------
ax -= np.mean(ax)
ay -= np.mean(ay)
az -= np.mean(az)
#-------------------------------------------------------------------------
# Estimate velocity:
#-------------------------------------------------------------------------
vx, vy, vz = velocity_from_acceleration(ax, ay, az, t)
#-------------------------------------------------------------------------
# Estimate position (dead reckoning):
#-------------------------------------------------------------------------
x, y, z, distance = position_from_velocity(vx, vy, vz, t)
print('distance = {0}'.format(distance))
return x, y, z, distance
def dead_reckon(ax, ay, az, t):
"""
Attempt dead reckoning (estimating position) from accelerometer data.
The accelerometer data are too noisy!
Parameters
----------
ax : list or numpy array of floats
accelerometer x-axis data
ay : list or numpy array of floats
accelerometer y-axis data
az : list or numpy array of floats
accelerometer z-axis data
t : list or numpy array of floats
accelerometer time points
Returns
-------
x : list or numpy array of floats
estimated position along x-axis
y : list or numpy array of floats
estimated position along y-axis
z : list or numpy array of floats
estimated position along z-axis
distance : float
estimated change in position
Examples
--------
>>> from mhealthx.extractors.dead_reckon import dead_reckon
>>> from mhealthx.xio import read_accel_json
>>> #input_file = '/Users/arno/DriveWork/mhealthx/mpower_sample_data/accel_walking_outbound.json.items-6dc4a144-55c3-4e6d-982c-19c7a701ca243282023468470322798.tmp'
>>> input_file = '/Users/arno/DriveWork/mhealthx/mpower_sample_data/deviceMotion_walking_outbound.json.items-5981e0a8-6481-41c8-b589-fa207bfd2ab38771455825726024828.tmp'
>>> start = 150
>>> device_motion = True
>>> t, axyz, gxyz, uxyz, rxyz, sample_rate, duration = read_accel_json(input_file, start, device_motion)
>>> ax, ay, az = axyz
>>> x, y, z, distance = dead_reckon(ax, ay, az, t)
"""
import numpy as np
from mhealthx.extractors.dead_reckon import velocity_from_acceleration,\
position_from_velocity
#-------------------------------------------------------------------------
# De-mean accelerometer readings:
#-------------------------------------------------------------------------
ax -= np.mean(ax)
ay -= np.mean(ay)
az -= np.mean(az)
#-------------------------------------------------------------------------
# Estimate velocity:
#-------------------------------------------------------------------------
vx, vy, vz = velocity_from_acceleration(ax, ay, az, t)
#-------------------------------------------------------------------------
# Estimate position (dead reckoning):
#-------------------------------------------------------------------------
x, y, z, distance = position_from_velocity(vx, vy, vz, t)
print('distance = {0}'.format(distance))
return x, y, z, distance
start = 150
t, axyz, gxyz, uxyz, rxyz, sample_rate, duration = read_accel_json(
input_file, start, device_motion)
ax, ay, az = axyz
#-------------------------------------------------------------------------
# De-mean accelerometer readings:
#-------------------------------------------------------------------------
ax -= np.mean(ax)
ay -= np.mean(ay)
az -= np.mean(az)
plotxyz(ax, ay, az, t, 'Acceleration')
#-------------------------------------------------------------------------
# Estimate velocity:
#-------------------------------------------------------------------------
vx, vy, vz = velocity_from_acceleration(ax, ay, az, t)
plotxyz(vx, vy, vz, t, 'Velocity')
#-------------------------------------------------------------------------
# Estimate position (dead reckoning):
#-------------------------------------------------------------------------
x, y, z, distance = position_from_velocity(vx, vy, vz, t)
print('distance = {0}'.format(distance))
plotxyz(x, y, z, t, 'Position')
#plotxyz3d(x, y, z)
#-------------------------------------------------------------------------
# De-mean accelerometer readings:
#-------------------------------------------------------------------------
ax -= np.mean(ax)
ay -= np.mean(ay)
az -= np.mean(az)
plotxyz(ax, ay, az, t, 'Acceleration')
#-------------------------------------------------------------------------
# Estimate velocity:
#-------------------------------------------------------------------------
vx, vy, vz = velocity_from_acceleration(ax, ay, az, t)
plotxyz(vx, vy, vz, t, 'Velocity')
#-------------------------------------------------------------------------
# Estimate position (dead reckoning):
#-------------------------------------------------------------------------
x, y, z, distance = position_from_velocity(vx, vy, vz, t)
print('distance = {0}'.format(distance))
plotxyz(x, y, z, t, 'Position')
#plotxyz3d(x, y, z)
