Maybe Matlab calculates gamma(X, a, b), not gamma (a, b, X)?

On 16/02/2016 13:08, Artur T. wrote:

Dear all,

I am just trying to replicate this MATLAB example of the Gamma
probability density function:

http://de.mathworks.com/help/stats/gampdf.html?requestedDomain=www.mathworks.com#zmw57dd0e292788

According to this, Matlab yields for shape=scale=1 and a x-vector seq(1,5) :
y = 0.3679  0.3033  0.2388  0.1947  0.1637

but with gretl I obtain correctly (just checked it with some statistical
tables)
y (1 x 5)
0.36788 0.13534 0.049787 0.018316 0.0067379

<hansl>
mu = seq(1,5)
y = pdf(g,1,1,mu)
</hansl>

Just out of curiosity: What is Matlab doing? (the same results are
obtained by Octave)

Artur