vis.frequency.dog_fit

  DOG_FIT - perform a difference of gaussians fit

  [PARAMS, ERR] = DOG_FIT(X,Y,S, ...)

  Perform a difference of gaussians fit Y_ for the equation

  Y_ = a1*exp(-X.^2/(2*b1^2)) - a2*exp(-X.^2/(2*b2^2))

  X and Y are column vectors with values of the fit. If S is provided,
  it is expected to be the standard deviation of the measurement and
  the fit will be weighted by 1/(1+s) for each entry (so more variable
  points get less weight).

  PARAMS are the parameters [A1 B1 A2 B2] of the DOG fit (see help DOG).
  ERR is the averaged squared error over all entries of Y, weighted
  by the weights calculated with S if provided.

  This function takes name/value pairs that influence its behavior
  |-----------------------------------------------------------------|
  | Parameter (default)  | Description                              |
  |----------------------|------------------------------------------|
  | start_positions (10) | Number of random start positions to use  |
  |_---------------------|------------------------------------------|