drcxcruz
commented
3 years ago Thanks for your interest in the module. I also had similar issue while testing. The module loads many library files in parallel. From your first screenshot, the library loading section is failing after line: using VegaLite; println("VegaLite")
Check that file GetPValue.jl is in a location that is found by the include Julia command. Google shows a number of posts about the "undefvarerror: include not defined" error. Apparently is not uncommon Julia issue. I believe you cannot simply change import and using.
Consider starting with new and blank Julia file. Then, add one line at time of the following to see which import is failing.
println("Loading Libraries... This can take a few minutes")
using Distributed;
using Hwloc; println("Hwloc")
if length(workers()) == 1
addprocs(Hwloc.num_physical_cores()) # get ready for parallelism
end
println("Distributed with ", length(workers()), " workers")
using DataFrames; println("DataFrames")
using Distributions: Normal, quantile; println("Distributions")
using Query: @rename, @orderby; println("Query")
using VegaLite; println("VegaLite")
include("GetPValue.jl"); println("GetPValue.jl")
include("GetBlankets.jl"); println("GetBlankets.jl")
using .GetBlankets: getBlanketAll, getBlanketBoosting, getBlanketRandomForest;
include("InvariantEnvironmentPrediction.jl");
using .InvariantEnvironmentPredictionModule; println("InvariantEnvironmentPrediction.jl")
include("ICPlibrary.jl"); println("ICPlibrary.jl")
println("Libraries Loaded")
Cheers
JuergenKanz
Your project is very interesting to me. Unfortunately, I can not run it. I have installed all needed packages like MLJ, DataFrames, CSV, Queryverse, and finally InvariantCausalPrediction on Julia 1.5.3.
In your Jupyter notebook ICP_Salary.ipynb I have changed the "include ...jl" to "using InvariantCausalPrediction".
Currently, I get the following output:
What can I do?