Issues

Issues with unconverted instances

There are instances that didn't get converted to .jl format due to a variety of reasons. Issues related to dimensionality, special operators, and others can be seen in the conversion process. As MINLPLibJuMP.jl develops, we are dedicated to resolve these issues and hope to improve the conversion rate on different libraries.

Issues with loading errors

With a converted .jl file, we test the instances by loading the JuMP model, initialize the NLPEvaluator objective, and collect some dimensional information from the loaded model. Given the scheme of model loading in JuMP environment is different from other modeling interfaces, we observed some exceptions and hogs during the process. Again, as MINLPLibJuMP.jl develops, we will dedicate to resolve these issues and hope to improve the overall performance in handling MINLP instances in JuMP.