Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# Re: st: unexpected -rbinomial- behaviour

 From Steve Samuels To statalist@hsphsun2.harvard.edu Subject Re: st: unexpected -rbinomial- behaviour Date Wed, 15 Sep 2010 11:25:19 -0400

--

" But it would have been easy enough to trap the endpoints and return
a meaningful (rather than missing) value."

Jeph, I don't think its so simple. What value would you report for
cases like this:

di rbinomial(1e+10,1e-9)    // n x p = 10
.
Steve

On Tue, Sep 14, 2010 at 5:26 PM, Jeph Herrin <stata@spandrel.net> wrote:
> But it would have been easy enough to trap the endpoints and
> return a meaningful (rather than missing) value.
>
>
> On 9/14/2010 4:34 PM, Tirthankar Chakravarty wrote:
>>
>> Jeph,
>>
>> The Stata function -rbinomial- is not defined for p=0. From
>> h rbinomial
>> rbinomial(n, p)
>>        Domain n:     1 to 1e+11
>>        Domain p:     1e-8 to 1-1e-8
>>        Range:        0 to n
>>
>> Your probabilistic statement about the degenerate Binomial
>> distribution is correct - the domain of the Binomial distribution is p
>> \in [0,1]. My guess would be that the p=/=0 condition is a limitation
>> of the simulation algorithm.
>>
>> T
>>
>> On Tue, Sep 14, 2010 at 12:54 PM, Jeph Herrin<stata@spandrel.net>  wrote:
>>>
>>> Am I wrong to expect rbinomial(n,0) = 0?
>>>
>>> . di rbinomial(10,0)
>>> .
>>>
>>> I would think that if P(success)= 0, then E(successes)=0.
>>>
>>>
>>> Jeph
>>>
>>>
>>> *
>>> *   For searches and help try:
>>> *   http://www.stata.com/help.cgi?search
>>> *   http://www.stata.com/support/statalist/faq
>>> *   http://www.ats.ucla.edu/stat/stata/
>>>
>>
>>
>>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/