generate random variables based on a non-standard t-distribution

[nathalie]

Hello all,


does anyone know how to generate random variables based on a non-standard t-distribution?
np.random only accepts the standard t distribution with df as input factor.

Best,
Nathalie

[ThomasL]

Hi, use Numpy.
https://docs.scipy.org/doc/numpy-1.14.0/...andom.html

[nathalie]

Hi Thomas, thanks for your reply.
Also numpy only takes the standard t distribution standard_t with (df[, size]) as input factors. However, I would like also input loc and scale.

[ThomasL]

So None of the 35 distributions you can choose from work for you?

[scidam]

does anyone know how to generate random variables based on a non-standard t-distribution?

Take a look at scipy.stats module. It allows to define custom distribution (continuous and discreet). What means non-standard t-distribution; maybe you can sample from standard t-dist and filter resulted points somehow.