MS_measure_trace_splitting

MS_MEASURE_TRACE_SPLITTING - Measure SWS for a pair of orthogonal traces

[fast, tlag] = MS_measure_trace_splitting(time, T00, T90, d_fast, ...
    d_tlag, max_tlag ...)

Inputs:
   time (vector) : Series of time points where amplitudes are known (s)
   T00 (vector)  : Series of amplitudes of transverse wave #1 before
                   splitting applied.
   T90 (vector)  : Series of amplitudes of transverse wave #2 before
                   splitting applied.
   d_fast (scalar) : Spacing of grid search for fast direction (degrees)
   d_tlag (scalar) : Spacing of grid search lag time (s)
   max_tlag (scalar) : Maximum lag time in grid search

Outputs:
   fast (vector)  : Optimum fast direction (degrees).
   tlag (vector)  : Optimum lag tim

Given a two traces (T00 and T90) on a common time axis (time) find
the splitting operator (fast and tlag) that minimises the second
eigenvalue of the covariance matrix. This is done in two steps. First
using an approximate grid search (spacing and time limits set by
d_fast, d_tlag and max_tlag) then by using the Matlab built
in simplex optimiser from the best grid point.

Options:

MS_measure_trace_splitting( … errorsurf) Draw a contour plot of the error surface generated during the grid search for the best splitting parameters.
MS_measure_trace_splitting( … optimoptions, options) Provide a structure of optimisation options to pass into the simplex optimiser. Use optimist to generate this. e.g. for looser tolerances use MS_measure_trace_splitting( … optimoptions, … optimset(TolFun, 1.0e-3, TolX, 1.0e-3)). Default options are: optimset(Display,none,TolFun,1.0e-5, … TolX, 1.0e-5).