Composite lognormal-pareto distribution

[silvia....@gmail.com]

Dear all, 

for the model I am working on, I would like to distribute a certain variable X as a composite lognormal-pareto. 

As a lognormal can be transformed from a normal distribution, I wrote in NetLogo: 

to-report  log-normal [mu sigma] 
 let beta ln (1 + ((sigma ^ 2) / (mu ^ 2))) 
 let M (ln(mu) - (beta / 2)) 
 let S (sqrt beta) 
 let x exp (random-normal M S) 
 report x 
end 

And I used the uniform distribution already built in in NetLogo to write a Pareto distribution: 

to-report pareto [alpha xmin] 
 report xmin / ( random-float 1 ^ (1 / alpha) ) 
end 

Now, how can I combine this in one composite distribution? Any suggestion? 

Thank you, 
Silvia