GENEActiv.calibrate {GENEAread}

GENEActiv.calibrate

Description

Function starts by identifying ten second windows of non-movement. Next, the average acceleration per axis per window is used to estimate calibration error (offset and scaling) per axis. The function provides recommended correction factors to address the calibration error and a summary of the callibration procedure.

Usage

GENEActiv.calibrate(
  binfile,
  use.temp = TRUE,
  spherecrit = 0.3,
  minloadcrit = 72,
  printsummary = TRUE,
  chunksize = c(),
  windowsizes = c(5, 900, 3600)
)

Arguments

binfile

A filename of a file to process.

use.temp

use temperature sensor data for calibration

spherecrit

The minimum required acceleration value (in g) on both sides of 0 g for each axis. Used to judge whether the sphere is sufficiently populated

minloadcrit

The minimum number of hours the code needs to read for the autocalibration procedure to be effective (only sensitive to multitudes of 12 hrs, other values will be ceiled). After loading these hours only extra data is loaded if calibration error has not been reduced to under 0.01 g

printsummary

if TRUE will print a summary when done

chunksize

number between 0.2 and 1 to specificy the size of chunks to be loaded as a fraction of a 12 hour period, e.g. 0.5 equals 6 hour chunks. The default is 1 (12 hrs). For machines with less than 4Gb of RAM memory a value below 1 is recommended.

windowsizes

Three values to indicate the lengths of the windows as in c(window1,window2,window3): window1 is the short epoch length in seconds and by default 5 this is the time window over which acceleration and angle metrics are calculated, window2 is the long epoch length in seconds for which non-wear and signal clipping are defined, default 900. However, window3 is the window length of data used for non-wear detection and by default 3600 seconds. So, when window3 is larger than window2 we use overlapping windows, while if window2 equals window3 non-wear periods are assessed by non-overlapping windows.

Details

The outputs from the function are as follows

• scale scaling correction values, e.g. c(1,1,1)

• offset offset correction values, e.g. c(0,0,0)

• tempoffset correction values related to temperature, e.g. c(0,0,0)

• cal.error.start absolute difference between Euclidean norm during all non-movement windows and 1 g before autocalibration

• cal.error.end absolute difference between Euclidean norm during all non-movement windows and 1 g after autocalibration

• spheredata average, standard deviation, Euclidean norm and temperature (if available) for all ten second non-movement windows as used for the autocalibration procedure

• npoints number of 10 second no-movement windows used to populate the sphere

• nhoursused number of hours of measurement data scanned to find the ten second time windows with no movement

• meantempcal mean temperature corresponding to the data as used for autocalibration.

Author(s)

Vincent T van Hees <[e-mail]> Zhou Fang Charles Sweetland <[e-mail]>

References

van Hees VT, Fang Z, Langford J, Assah F, Mohammad A, da Silva IC, Trenell MI, White T, Wareham NJ, Brage S. Auto-calibration of accelerometer data for free-living physical activity assessment using local gravity and temperature: an evaluation on four continents. J Appl Physiol (1985). 2014 Aug 7