I was wondering whether or not is there a way to fit my data to a power law with JAGS.... something on the lines of
y_ ~ dPowerLaw(a,b)
assuming that the Power Law distribution will be specified by y=a*x^b
Thanks.
_
You want the Pareto distribution. It is described in the user manual.
I'm not sure, it seems that the Pareto distribution would assumes some relationship between (what i called) "a" and "b".
a = alpha * c^alpha
b = -(alpha+1)
I'm would like to avoid that.
What you call "a" is the normalizing constant. It is not a free parameter but is determined by the fact that the density must integrate to 1.
Log in to post a comment.
I was wondering whether or not is there a way to fit my data to a power law
with JAGS.... something on the lines of
y_ ~ dPowerLaw(a,b)
assuming that the Power Law distribution will be specified by y=a*x^b
Thanks.
_
You want the Pareto distribution. It is described in the user manual.
I'm not sure, it seems that the Pareto distribution would assumes some
relationship between (what i called) "a" and "b".
a = alpha * c^alpha
b = -(alpha+1)
I'm would like to avoid that.
What you call "a" is the normalizing constant. It is not a free parameter but
is determined by the fact that the density must integrate to 1.