Hello Georgiana. We've moved a lot of the functionality provided by Rmath out into separate packages, such as Distributions.jl, which I believe should host the functionality you're looking for, although I think it's under a different name than `dnorm`. I think something like the following should give you dnorm-like functionality:
julia> using Distributions
julia> µ = 1.0
1.0
julia> σ = 1.5
1.5
julia> d = Normal(µ, σ)
Normal(μ=1.0, σ=1.5)
julia> pdf(d,1.0)
0.2659615202676218
Note that you will need to have installed the Distributions package for this to work, do that via Pkg.add("Distributions"). If for some other reason you really need an older version of Julia, we do host them, but the links aren't published anywhere at the moment. Here are the links for your convenience, although I do highly suggest massaging your script to work on the 0.3.X line of Julia releases; there are many, many improvements to usability, speed, and language coherence. Also note that we cannot provide support for this old of a release of Julia, if you run into bugs or such, the most likely response to any bug report will be "please upgrade to 0.3.5".
0.2.1 download links:
If you are on Linux, you may be able to install Julia 0.2.1 through your package manager by explicitly asking for an older version. You may also be able to compile Julia from source, although many of the dependencies probably have newer versions released and may not be distributing their sources at the same URLs the buildscripts will attempt to download them from.
-E