First of all, I'd like to express my appreciation for the very nice preconditioners! They are almost magic, bringing down the number of iterations required to solve Poisson's problem from almost maxiter to about 5!
I discovered the following error when testing my package on AppVeyor, which also tests on Win32:
MethodError: no method matching AlgebraicMultigrid.Solver(::AlgebraicMultigrid.Classical{Float64}, ::AlgebraicMultigrid.RS, ::AlgebraicMultigrid.GaussSeidel{AlgebraicMultigrid.SymmetricSweep}, ::AlgebraicMultigrid.GaussSeidel{AlgebraicMultigrid.SymmetricSweep}, ::Int32, ::Int32)
Closest candidates are:
AlgebraicMultigrid.Solver(::S, ::T, ::P, ::PS, !Matched::Int64, !Matched::Int64) where {S, T, P, PS} at C:\Users\appveyor\.julia\packages\AlgebraicMultigrid\xJj9G\src\classical.jl:2
First of all, I'd like to express my appreciation for the very nice preconditioners! They are almost magic, bringing down the number of iterations required to solve Poisson's problem from almost
maxiter
to about 5!I discovered the following error when testing my package on AppVeyor, which also tests on Win32:
The reason is that
max_levels
andmax_coarse
are hard-coded to beInt64
: https://github.com/JuliaLinearAlgebra/AlgebraicMultigrid.jl/blob/master/src/classical.jl#L6 Shouldn'tInt
be enough?