Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I have the expression

$$\displaystyle\frac{\beta(x + a, y + b)}{\beta(a, b)}$$

where $\beta(a_1,a_2) = \displaystyle\frac{\Gamma(a_1)\Gamma(a_2)}{\Gamma(a_1+a_2)}$.

I have a feeling this should have a closed-form which is intuitive and makes less heavy use of the Beta function. Can someone describe to me whether this is true?

Here, $x$ and $y$ are integers larger than $0.$

share|improve this question

migrated from stats.stackexchange.com Aug 17 '12 at 22:18

This question came from our site for people interested in statistics, machine learning, data analysis, data mining, and data visualization.

    
I don't see any distributions in this question. Are you just asking for a simplification of this ratio of values of the Beta function? Are there any restrictions on $x$ and $y$ (possibly they are integers? Non-negative numbers? Real numbers)? –  whuber Aug 17 '12 at 21:41
    
sorry, yes, this question originates in a problem related to the Beta distribution, maybe I should have mentioned that explicitly. $x$ and $y$ are indeed integers. –  singelton Aug 17 '12 at 21:59
    
After typing this into wolfram alpha I would say the short answer is "no". –  martin Aug 17 '12 at 23:15
    
The beta function is written with a capital beta $B(x+a,y+b)/B(a,b)$ –  Américo Tavares Aug 18 '12 at 10:58

3 Answers 3

$$ \beta(1+a,b) = \frac{\Gamma(1+a)\Gamma(b)}{\Gamma(1+a+b)} = \frac{a\Gamma(a)\Gamma(b)}{(a+b)\Gamma(a+b)} = \frac{a}{a+b} \frac{\Gamma(a)\Gamma(b)}{\Gamma(a+b)} = \frac{a}{a+b} \beta(a,b). $$ If you have, for example $\beta(5+a,8+b)$, just repeat this five times for the first argument and eight for the second: $$ \frac{(4+a)(3+a)(2+a)(1+a)\cdot(7+b)(6+b)\cdots (1+b)b}{(12+a+b)(11+a+b)\cdots (1+a+b)(a+b)}\beta(a,b). $$

share|improve this answer

I think Michael Hardy gives your answer. I only want to notify that if $y=0$ then

$x$th raw moment of Beta distribution $\mu_x^'=E(T^x)=\frac{\int_0^1 t^{x+a-1}(1-t)^{b-1}~dt}{\beta(a,b)}=\frac{\beta(a+x,b)}{\beta(a,b)}$ if $x>-a$

share|improve this answer

You can also write this as a ratio of rising factorials. The rising factorial is defined as $(a)_k = a(a+1)\cdots(a+k-1)$, with $(a)_0 = 1$. Then use the recurrence relationship for the Gamma function to reduce the ratio of the beta functions to $$\frac{\beta(a+x,b+y)}{\beta(a,b)} = \frac{(a)_x(b)_y}{(a+b)_{(x+y)}}.$$ I think that's as concise as it gets. Of course, some folks aren't too fond of rising factorial notation...

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.