Skip to contents

Integrate EMG and MoCap signals onto a common timeline

Source: R/mocap-emg.R
integrateEMGMoCap.Rd

Processes EMG at its native sampling rate, then aligns the resulting envelope (or MVC-normalized envelope) to MoCap sample times for feature analysis. Start times must be expressed on a shared clock. Alignment uses linear interpolation with the selected support policy; it does not estimate offsets. With process = FALSE, raw samples are interpolated without anti-alias filtering. With outside = "NA", combined features remain missing outside raw-data support even if processing produces finite values beyond that support.

Usage

integrateEMGMoCap(
  mocap,
  emg,
  mocap_sampling_rate = NULL,
  emg_sampling_rate,
  mocap_assay = NULL,
  process = TRUE,
  ...,
  emg_start_time = 0,
  mocap_start_time = 0,
  outside = c("extend", "NA", "error")
)

Arguments

mocap

A numeric matrix/data.frame (time x features), or a PhysioExperiment object.

emg

Numeric vector or matrix (time x channels).

mocap_sampling_rate

MoCap sampling rate in Hz. If mocap is a PhysioExperiment and this is NULL, uses samplingRate(mocap).

emg_sampling_rate

EMG sampling rate in Hz.

mocap_assay

Assay name to use when mocap is a PhysioExperiment.

process

Logical; if TRUE, runs processEMG() before combining.

...

Additional arguments passed to processEMG().

emg_start_time, mocap_start_time

Finite scalar start times in seconds on the same reference clock. Defaults assume both recordings start at zero. These are known offsets; no synchronization or clock drift is estimated.

outside

Handling of target times outside each channel's finite-data support: "extend" repeats endpoints (compatibility default), "NA" leaves unavailable times missing, and "error" rejects the alignment. Fewer than two finite observations produce missing output, or an error under "error". Internal gaps are still interpolated; this policy checks support boundaries, not internal gap duration. No anti-alias filter is added. Endpoint differences within floating-point roundoff are snapped to the endpoint (tolerance capped at one millionth of the shorter sample interval).

Value

A list with mocap, emg_aligned (aligned raw EMG, retained for compatibility), and combined data.frame (processed EMG when requested).

References

Merletti R, Parker PA (2004). "Electromyography: Physiology, Engineering, and Non-Invasive Applications." IEEE Press/Wiley.

See also

processEMG() for EMG processing pipeline, alignEMGtoMoCap() for time-alignment of EMG to MoCap, synchronizeSignals() for general multi-signal synchronization.

Examples

mocap <- matrix(rnorm(500), ncol = 5)
emg <- matrix(rnorm(5000), ncol = 2)
out <- integrateEMGMoCap(mocap, emg, mocap_sampling_rate = 100,
                         emg_sampling_rate = 1000)