add test for nan derivatives for distribution tests

[syclik]

the distribution tests currently don't check if the derivatives are nan. I think they should.

[syclik]

I fixed the distribution tests. I located the distributions that have problems. We'll need to fix all of them by hand.

Warnings:

• neg_binomial

Failed tests:

• beta
• cauchy
• double_exponential
• exp_mod_normal
• exponential
• frechet
• gumbel
• lognormal
• neg_binomial_2
• normal
• pareto_type_2
• rayleigh
• skew_normal
• weibull

[bob-carpenter]

Awesome --- this should do wonders for robustness.

Is the plan to have them throw if they catch NaN values?
I think that'd be better than returning NaN values, which
would cause rejection, but then would fail in the double
case where you didn't care about gradients.

• Bob

On Mar 17, 2015, at 8:28 AM, Daniel Lee notifications@github.com wrote:

I fixed the distribution tests. I located the distributions that have problems. We'll need to fix all of them by hand.

Warnings:

• neg_binomial
Failed tests:

• beta
• cauchy
• double_exponential
• exp_mod_normal
• exponential
• frechet
• gumbel
• lognormal
• neg_binomial_2
• normal
• pareto_type_2
• rayleigh
• skew_normal
• weibull
—
Reply to this email directly or view it on GitHub.