checkEstimates

checkEstimates(ss_est, ss_mdl, y_true, x_true, alpha_true, ss_gen, K_lim, T_lim)

checkEstimates takes EKF/IEKF estimates of a multi-peak state-space model applied to simulated data and compares them to the true values to evaluate the filter performance based on a set of statistical measures:

It uses several internal functions given below:

icTest, Lmatr, Dmatr, combo2PickParams, combo2PickPeaks, getComboNum

INPUTS:

ss_est – estimate struct containing filter estimate variables.

Required.

ss_mdl – StateSpaceMultiPeak object used to obtain filter

estimates. Required.

y_true – matrix of true observations (dim_y x NT). Required. x_true – matrix of true states (dim_x x NT+1). Required. alpha_true – indicator matrix of true On/Off-combinations

(num_combos x NT+1). Required.

ss_gen – StateSpaceMultiPeak object used to generate the true

observations. Required.

K_lim – Positive integer sets number of frequency bins over which

correlations are evaluated in icTest, boxQ, and moranI. Default 3.

T_lim – Positive integer sets number of time steps over which

correlations are evaluated in boxQ and moranI. Default 4.

OUTPUTS:

measures – a vector containing statistical values (15 x 1)

Created by Patrick Stokes