PyPSA/EnergyModels.jl

Update to Julia 1.1 and JuMP 0.19 (ie. MOI)

Closed this issue · 2 comments

coroa commented
  • Julia 1.1 code
  • Replace AxisArray with JuMP's DenseAxisArray
  • Re-do @emvariable, @emconstraint macros based on macrocall
coroa commented

DenseAxisArray does not have names for its axes, so that WrappedArray cannot be used to augment the Array automatically. On the other hand will the compilation overhead probably decrease since the type information does change less often :).

Three possible solutions:

  1. The current code based AxisArrays works again, with a small one line modification that makes Axis types iterable again (JuliaArrays/AxisArrays.jl#160).

  2. We could easily push the axes augmentation into get(::Data, ...), where the axes info is still known from the NetCDF itself.

coroa commented

Fixed by 31d13df.