Rev Language Reference

dnBimodalLognormal(Real mean1, Real mean2, RealPos sd1, RealPos sd2, Probability p)

mean1 :	Real (pass by const reference)
	The mean (in log-space) of the first lognormal distribution.
mean2 :	Real (pass by const reference)
	The mean (in log-space) of the second lognormal distribution.
sd1 :	RealPos (pass by const reference)
	The standard deviation (in log-space) of the first lognormal distribution.
sd2 :	RealPos (pass by const reference)
	The standard deviation (in log-space) of the second lognormal distribution.
p :	Probability (pass by const reference)
	The probability to belong to the first distribution.

RealPos

The bimodal lognormal distribution takes five parameters: mean1 and mean2 (the means (in log-space) of the two lognormal distributions), sd1 and sd2 (the standard deviations (in log-space) of the two lognormal distributions), and p (between 0 and 1). The value will be distributed according to the first lognormal distribution with probability p, and distributed according to the second lognormal distribution with probability (1 - p).

p ~ dnBeta(1.0,1.0)
x ~ dnBimodalLognormal(mean1=-1,mean2=1,sd1=0.1,sd2=0.1,p=p)
x.clamp( exp(1) )
moves[1] = mvSlide(p, delta=0.1, weight=1.0)
monitors[1] = screenmonitor(printgen=1000, separator = "        ", x)
mymodel = model(p)
mymcmc = mcmc(mymodel, monitors, moves)
mymcmc.burnin(generations=20000,tuningInterval=100)
mymcmc.run(generations=200000)

• dnBimodalNormal
• dnLognormal