Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[WIP] Fractal with bconds rcnstrct from 2dx #492

Draft
wants to merge 134 commits into
base: master
Choose a base branch
from
Draft

[WIP] Fractal with bconds rcnstrct from 2dx #492

wants to merge 134 commits into from

Conversation

pdziekan
Copy link
Contributor

Like fractal reconstruction with bcnds, but here reconstruction is done not from the scales of dx, but from 2*dx, i.e. every second resolved point.
NOTES:

  • each thread needs at least 5 points along y axis, compared to 3 in reconstruction from dx (this could be fixed by making some threads not reconstructing at all, it's a matter of refinemenet_ranges)
  • in PBL this approach gives more (too much?) energy in smallest scales
  • in UWLCM this gives larger spread of th and rv within cloud

@pdziekan
sgs_fra solver declaration
c5a37c0
@pdziekan
wip on fractal refinement
f915875
@pdziekan
working allocation of refined arrays
c1b497d
@pdziekan
refined grid size dependant on the number of fractal reconstructions
487b356
@pdziekan
fix refine_grid_size()
1ad25f3
@pdziekan
mem_factory function
f3885aa
@pdziekan
consitional choice of shmem / shmem_refined
47348d7
@pdziekan
consitional choice of shmem / shmem_refined - upgraded
a6e8b43
@pdziekan
working shmem / shmem_recon stuff? (openmp and 3d only for now)
be93f45
@pdziekan
solver_family protected again + solver doesnt inherit from family tag
963be45
@pdziekan
use mem_factory in all concurr types
d361919
@pdziekan
second ctor for shmem_refined in mem_t in each concurr type
2d956d1
@pdziekan
alloc tmp with custom grid size in 1d and 2d
7d54514
@pdziekan
hdf5: store refined data
3f81179
@pdziekan
xmf: record refined aux
71f4d70
@pdziekan
hdf5: store refined X,Y,Z TODO: make it conditional
0fe0730
@pdziekan
xmf writer: add reference writing
caaffb4
@pdziekan
xmf: write refined data using separate xmf_writer
1beeef1
@pdziekan
Merge branch 'master' of github.com:igfuw/libmpdataxx into fractal
cb7d541
@pdziekan
fix grid size refinement strategy
54e392b
@pdziekan
wip on refinement: fixed positions in output, added refined ijk
fb08f52
@pdziekan
refined fields without halos
e9984b5
@pdziekan
refined grid: halo of n_ref/2
3f540a9
@pdziekan
Revert "refined grid: halo of n_ref/2"
87df456 
This reverts commit 3f540a9.
@pdziekan
wip on refined interpolation
aba74b3
@pdziekan
wip on interpolation
6aab97c
@pdziekan
interpolation: hardocode two iterations, still no interpolation at sh…
e622df1 
…aredmem and MPI boundaries
@pdziekan
wip on interpolation at edges
669194a
@pdziekan
shmem_ref: require grid size that makes fractal reconstruction easier
fe9d8f8
@pdziekan
refinement: wip
275789c
pdziekan and others added 29 commits March 24, 2023 17:47
@pdziekan
record aux dsc hlpr ref/nonref separation: compiling version (3D only)
d0e4f64
@pdziekan
ref hdf5 fixes
b8ed801
fill refinee distmem only if distmem size > 1
38c72db
@pdziekan
xchng_ref after generating d_j
11579de
use halo in spatial_avergae_ref_2_reg; TODO: if it works, grid_size i…
ff8c2a4 
…s not needed
Revert "use halo in spatial_avergae_ref_2_reg; TODO: if it works, gri…
8570037 
…d_size is not needed"

This reverts commit ff8c2a4.
spatial average r2r weighting: account for domain edges
6ca6059
add avg_edge_sclr_ref and spatial_average_ref2reg
ee3a1f4
avg_edge_sclr_ref and spatial_average protected
d35ae71
remove nancheck
dd89289
DBG: dont do spatial averaging ref_2_reg, just copy corresponding cells
72f8051
Revert "DBG: dont do spatial averaging ref_2_reg, just copy correspon…
a11ee8e 
…ding cells"

This reverts commit 72f8051.
Revert "Revert "DBG: dont do spatial averaging ref_2_reg, just copy c…
d9d299c 
…orresponding cells""

This reverts commit a11ee8e.
do avg_edge_sclr_ref after refinemenet/interpolation
8565eb0
more barriers in refine/interpolate
f0e93f0
full spatial_ref_2_reg
08a30e6
Revert "full spatial_ref_2_reg"
312af4f 
This reverts commit 08a30e6.
Revert "Revert "full spatial_ref_2_reg""
2d29583 
This reverts commit 312af4f.
@pdziekan
stretching parameters from E. Akinlabi: move the to separate namespace
8a27891
@pdziekan
for reference, include the stretching parameter data from E. Akinlabi
753f46e
@pdziekan
include the stretching parameter data from M. Waclawczyk
38dc714
@pdziekan
add a readme to stretching parameters from Marta
392867c
@pdziekan
stretchin parameters from LES data: formulae for th subsat
c9588ae
@pdziekan
formulae: move stretchng parameters to separate file
b8deacd
@pdziekan
other stretching parameters WIP: check distribution of values, seems …
d7ca208 
…to bo a lot of extremes
@pdziekan
Marta's stretch params: dat->hpp
c5be8a3
@pdziekan
add other stretching parameters from Marta
00e5135
@pdziekan
double stride of ijk_r2r and require nx/ny/nz=5+4*i
d6f5349
@pdziekan
reconstruct from every second point
593aee8
@pdziekan pdziekan changed the title Fractal with bconds rcnstrct from 2dx [WIP] Fractal with bconds rcnstrct from 2dx May 12, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
1 participant
@pdziekan