Julia Ecosystems, Packages and panel data

question

#1

I have a few questions as a new user. 1. What is a Julia ecosystem? I am guessing it is something to do with the version of the language.

I am trying to construct a regression with panel data and was told by someone to use the following (for a fixed effects model):

data = CSV.read(“wagepan.csv”);
categorical!(data, [:nr, :year, :married, :black, :hisp, :educ])
fm = @formula lwage ~ exper + expersq + married + union + (absorb = nr + year)
model = EconometricsModel(fm, data)
coeftable(model)

and they mentioned that the reason mine is not working is because I am using JuliaPro hence my first question. Is there a way to use JuliaPro to run the “categorical!” function? and if I were to run the above in JuliaPro what packages will I need.

Thanks,
J


#2

Where are you seeing the term? Hard to say otherwise.

What’s the context here? Is it because JuliaPro doesn’t ship with whatever package provides EconometricsModel? Or perhaps because the versions of packages that ship with JuliaPro aren’t recent enough? Or are too new?


#3

Are you coming from Matlab? It sounds like you’re used to the toolboxes built into Matlab.

Most functionality in Julia is in packages. For fixed effects, I think https://github.com/matthieugomez/FixedEffectModels.jl is the best right now,
You can use it by typing

Pkg.add("FixedEffectModels")
using FixedEffectModels

and then following the documentation.

I remember @Nosferican said he was developing a panel data package with EconometricsModel, but I don’t know if it’s available at the moment.


#4

Generally, when people talk about a Julia ecosystem, they are talking about a set of related packages, just as the ones that work with Plots.jl, or the whole set of packages in the JuliaDiffEq organization.
(similar to the “Hadleyverse” in R)
There really isn’t anything particularly specific to any version of Julia (although you may need some minimal version of Julia to use the packages in a particular ecosystem, some developers are dropping v0.5 and earlier support)


#5

Econometrics.jl is mostly coded and has panel data estimators (pooling OLS, between, within, and random effects Swamy-Arora harmonic mean). However, I am holding up until 0.7-rc/1.y.z and the JuliaStats ecosystem to release the package. For Julia 0.6.z, you could clone JuliaEconometrics/EconUtils.jl and use the partial demeaning for Random Effects, but it isn’t documented yet.