Skip to main content
. Author manuscript; available in PMC: 2014 Oct 9.
Published in final edited form as: J Comput Neurosci. 2013 Sep 6;36(3):355–382. doi: 10.1007/s10827-013-0477-1

Fig. 3.

Fig. 3

Error decomposition. Dark arrows show the decomposition of vector error into channels implemented in the gaze cerebellar controller. G0 corresponds to the initial gaze position. GF is the desired final gaze position. εΔG corresponds to the current gaze error. εΔG,‖ is the projection of the error along the desired initial gaze displacement (the vector from G0 to GF). This corresponds to the remaining gaze displacement along the desired displacement (dashed line). εΔG,⊥ corresponds to the perturbation with respect to the desired gaze displacement. A similar decomposition has been implemented for the control of the head trajectory