GridapROMs.Distributed

module Distributed

MPI-parallel extensions of the GridapROMs parametric and reduced-basis layer.

Wraps the distributed data structures provided by PartitionedArrays and GridapDistributed so that parametric snapshots and FE solves can be performed in parallel. Key components:

• Distributed ParamArrays — OwnAndGhostParamVector and ParamJaggedArray carry multi-sample DOF arrays in the owned-and-ghost partitioning expected by PartitionedArrays. ParamArraysInterface.jl implements the full AbstractParamArray interface for these types.

• Sparse utilities — ParamSparseUtils.jl provides distributed sparse matrix/vector assembly helpers (CSR row-pointer arithmetic, assembly caches) compatible with ParamArray entries.

• Primitives — Primitives.jl contains low-level MPI-aware operations (scatter/gather, consistent local-size queries) reused by the higher-level components.

• Distributed algebra — ParamAlgebra.jl extends ParamBuilder / ParamCounter to the distributed setting; ParamSolvers.jl wraps GridapDistributed solvers for parametric systems.

• Distributed FE spaces — ParamFESpaces.jl specialises DistributedSingleFieldFESpace / DistributedMultiFieldFESpace for parametric DOF arrays.

• Distributed snapshots — GenericPArray is a PartitionedArrays PVector-like container for generic parallel data; DistributedSnapshots wraps Snapshots for the distributed case.

Requires GridapDistributed and MPI; not loaded unless Distributed is explicitly used.

struct GenericPArray{V,A,B,C,D,T,N} <: AbstractArray{T,N}
  array_partition::A
  index_partition::B
  unpartitioned_axes::C
  cache::D
end

Same as PVector, but while the latter always stores a vector with entries partitioned on different cores, this structure stores an array (not necessarily a vector) partitioned along the first dimension (row-wise)

GenericPArray(array_partition,index_partition)

Create an instance of GenericPArray from the underlying properties array_partition and index_partition.

struct OwnAndGhostParamVectors{A,B,T} <: AbstractParamVector{T}
  own_values::A
  ghost_values::A
  permutation::B
end

Parametric version of OwnAndGhostVectors