DoG, fit first derivative of Gaussian

[ALPHA,SIGMA, AMP] = DOG(X,Y) fits first derivative of Gaussian to
x,y-data by minimizing the sum of squared residuals. The output parameter
ALPHA controls amplitude and SIGMA is the standard deviation of the
Gaussian distribution and controls width of the resulting curve, given by
y = normpdf(x,0,SIGMA).*x.*ALPHA. AMP is the peak amplitude.

Is often used in research on serial dependence, e.g. doi:10.1038/nn.3689