vlt.neuro.vision.speed.fit

  vlt.neuro.speed.fit Fit two-dimensional Gaussian function to data set

   P = vlt.neuro.vision.speed.fit(SF,TF,R, min_xi)

   Fit a Gaussian to a set of responses generated by a speed tuning curve.

   Inputs:
     SF is an array of spatial frequency values
     TF is an array of temporal frequency values
     R is an array of measured responses driven by the spatial and
     temporal frequency values
     MIN_XI: if provided, provides the lower limit on XI. If not provided,
     assumed to be 0. (Might provide -1, for example.)

   Outputs:
     P is an array with parameters:
     ---------------------------------------------
     | A         | Peak response of the neuron   |
     |           |                               |
     | zeta      | Skew of the temporal          |
     |           | frequency tuning curve        |
     |           |                               |
     | xi        | Speed parameter               |
     |           |                               |
     | sigma_sf  | Tuning width of the neuron    |
     |           | for spatial frequency         |
     |           |                               |
     | sigma_tf  | Tuning width of the neuron    |
     |           | for temporal frequency        |
     |           |                               |
     | sf0       | Preferred spatial frequency   |
     |           | averaged across temporal      |
     |           | frequencies                   |
     |           |                               |
     | tf0       | Preferred temporal frequency  |
     |           | averaged across spatial       |
     |           | frequencies                   |
     ---------------------------------------------

   See: Priebe et al. 2006

   By Noah Lasky-Nielson