Re-implement ZeroMeanFESpace

NEWS.md

@@ -5,7 +5,12 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
-## Unreleased
+## [0.4.4] 2024-08-14
+
+### Added
+
+- Added kwargs for VTK encoding options. Since PR[#156](https://github.com/gridap/GridapDistributed.jl/pull/156).
+- Reimplemented distributed ZeroMeanFESpaces. Since PR[#155](https://github.com/gridap/GridapDistributed.jl/pull/155).
 
 ### Fixed
 

Project.toml

@@ -1,7 +1,7 @@
 name = "GridapDistributed"
 uuid = "f9701e48-63b3-45aa-9a63-9bc6c271f355"
 authors = ["S. Badia <santiago.badia@monash.edu>", "A. F. Martin <alberto.f.martin@anu.edu.au>", "F. Verdugo <f.verdugo.rojano@vu.nl>"]
-version = "0.4.3"
+version = "0.4.4"
 
 [deps]
 BlockArrays = "8e7c35d0-a365-5155-bbbb-fb81a777f24e"

src/BlockPartitionedArrays.jl

@@ -25,6 +25,21 @@ function Base.getindex(a::BlockPRange,inds::Block{1})
   a.ranges[inds.n...]
 end
 
+function PartitionedArrays.matching_local_indices(a::BlockPRange,b::BlockPRange)
+  c = map(PartitionedArrays.matching_local_indices,blocks(a),blocks(b))
+  reduce(&,c,init=true)
+end
+
+function PartitionedArrays.matching_own_indices(a::BlockPRange,b::BlockPRange)
+  c = map(PartitionedArrays.matching_own_indices,blocks(a),blocks(b))
+  reduce(&,c,init=true)
+end
+
+function PartitionedArrays.matching_ghost_indices(a::BlockPRange,b::BlockPRange)
+  c = map(PartitionedArrays.matching_ghost_indices,blocks(a),blocks(b))
+  reduce(&,c,init=true)
+end
+
 """
   struct BlockPArray{V,T,N,A,B} <: BlockArrays.AbstractBlockArray{T,N}
 """
@@ -336,6 +351,11 @@ function LinearAlgebra.dot(x::BlockPVector,y::BlockPVector)
 end
 
 function LinearAlgebra.norm(v::BlockPVector,p::Real=2)
+  if p == 2
+    # More accurate, I think, given the fact we are not 
+    # repeating the sqrt(square(sqrt...)) process in every block and every processor
+    return sqrt(dot(v,v)) 
+  end
   block_norms = map(vi->norm(vi,p),blocks(v))
   return sum(block_norms.^p)^(1/p)
 end