MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am sampling the distribution of lifetimes of computers participating in massive volunteer computing initiatives (BOINC projects). While a phenomenological Weibull distribution makes a good description, I find that a better fit happens with a Gamma distribution with a shape parameter $k < 1$

Now, while a Gamma distribution with integer shape parameter is produced as the waiting time for the $k$-th event of a Poisson process, I do not know of a similar, process-based, description of the meaning of a Gamma distribution for a non integer parameter, nor to say for non integer and less than unity. Is there one?

share|cite|improve this question
-- it could be possible to think of the lifetime as a difference between two random variables: the registered user decides to incorporate a computer to the project, and the registered user decides to extract a computer out of the project. But this combination could be a red herring; obviously the time of registration of the first computer is very near of the time of registration of the user. -- the measured shape parameter is about $0.26...$ so that the gamma distribution has a term going as $x^{-0.73...}$ near zero. – arivero Sep 15 '10 at 14:05

This is just a hint (actually yours is a question I planned to ask here myself sooner or later): What's the model for the gamma distribution with non-integer parameter $\lambda$?

Since possibly the most relevant aspect is the semigroup law convolution, my feeling is that it should describe properly waiting times for reaching given amount $\lambda$ of mass, arriving continuously and randomly under independence assumptions, like e.g. rain, falls of powder on the soil. I'm very interested in further more precise answers.

share|cite|improve this answer
There is a severe problem with any stationary mass accumulation model with independent increments. Since the density tends to infinity at $0$, we get that the probability $p(t)$ to get the critical mass over every interval of small length $t>0$ is much bigger than $t$. But then the probability not to get it over any positive time $T$ is at most $(1-p(t))^{T/t}\to 0$ as $t\to 0$. This doesn't mean that no rainfall model is possible, just that it has to have a feature that big rains can occur but they should exhaust the water supply and make the subsequent rains small with high probability. – fedja Sep 15 '10 at 15:30

Here is some interesting, though natural, view of the problem (including a small but quite elegant result at the end). Denote by ${\rm Gamma}(k,\lambda)$ the gamma distribution with density function $f(x;k,\lambda) = \lambda ^k x^{k - 1} e^{ - \lambda x}/ \Gamma (k)$, $x>0$. Here $k$ and $\lambda$ are arbitrary positive constants. Clearly, the problem reduces to the case $0<k<1$. Indeed, if $k$ is non-integer and greater than $1$, then a ${\rm Gamma}(k,\lambda)$ random variable can be viewed as a sum of independent ${\rm Gamma}(\left\lfloor k \right\rfloor,\lambda)$ and ${\rm Gamma}(k - \left\lfloor k \right\rfloor,\lambda)$ random variables. Now, let's consider a gamma process $X = \lbrace X_t : t \geq 0\rbrace$ such that $X_1 \sim {\rm exponential}(\lambda)$. That is, $X$ is a process with stationary independent increments, starting at $0$ and having right-continuous with left limits sample paths (i.e., a L\'evy process), such that $X_t - X_s \sim {\rm Gamma}(t-s,\lambda)$, for any $0 \leq s < t$. The process $X$ is a pure jump process, having positive jumps only, and it holds $X_t = \sum\nolimits_{0 < s \le t} {\Delta X_s } $, where $\Delta X_s = X_s - X_{s-}$. The number of jumps is a.s. infinite countable in any time interval, no matter how small (yet, $\Delta X_t =0$ a.s., for any fixed $t>0$). The process $X$ is characterized by its jump measure (L\'evy measure) given by $\nu({\rm d}x)=x^{-1}e^{-\lambda x}\,{\rm d}x$: given a time interval $[t_1,t_2]$, the number of jumps of $X$ whose sizes lie in $[a,b] \subset (0,\infty)$ is Poisson distributed with mean $(t_2-t_1)\int_a^b {x^{ - 1} e^{ - \lambda x } \,{\rm d}x}$ (the key concept here is Poisson random measure). Hence, roughly speaking, the process usually increases by tiny jumps (whose sum is negligible in practice).

Now, consider the process $X$ on the time interval $[0,1]$. On the one hand, $X_1$ is the (infinite) sum of jumps of $X$ up to time $t=1$. On the other hand, $X_1$ is distributed as the waiting time until the first occurrence in a Poisson process with rate $\lambda$. This is already quite interesting. Returning to the original problem, we can view a ${\rm Gamma}(k,\lambda)$ random variable, $0<k<1$ fixed, as the random variable $X_k$, i.e., the sum of jumps of $X$ up to time $k$ (note that $\lim _{k \uparrow 1} X_k = X_1 $, a.s.). So, in some respect, this already gives a process-based description of the meaning of a Gamma distribution for a non-integer parameter. But much more can be said, using the fact that if $X$ and $Y$ are independent ${\rm Gamma}(\alpha,\lambda)$ and ${\rm Gamma}(\beta,\lambda)$ rv's, respectively, then the ratio $X/(X+Y)$ has a ${\rm Beta}(\alpha,\beta)$ distribution. Namely, the ratio $R = X_k / X_1$, $0<k<1$, is distributed as a ${\rm Beta}(k,1-k)$ random variable, implying that $R$ has density function $f_{R} (x;k) = x^{k-1}(1-x)^{-k}\sin(\pi k)/\pi$, $0<x<1$, where we have used $\Gamma(k)\Gamma(1-k) = \pi/\sin(\pi k)$. In particular, $f_R (x;1/2) = 1/(\pi \sqrt {x(1 - x)} )$, $0<x<1$.

share|cite|improve this answer

Your Answer


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

Not the answer you're looking for? Browse other questions tagged or ask your own question.