Lots of stuff... #125

KristofferC · 2017-03-20T16:07:51Z

No description provided.

fredrikekre · 2017-03-20T16:24:45Z

Nice, possible to extract renaming commits to another PR and merge that?

KristofferC · 2017-03-20T16:28:59Z

Yes

add face iterator remove the redundant face list fixup iterators remove stupid getface

update for file move

KristofferC · 2017-03-23T13:47:33Z

Sorry for the mess... I will try to fix it up later. The interesting changes right now are the following:

Fix so looping over faces is more natural
Add a DofHandler type
Add a DirichletBoundaryCondition type.
Add a create_sparsity_pattern function that creates the sparsity pattern from a dofhandler.
Add a new assembler type that takes a sparse matrix and efficiently assemble into it.
Create a function apply! that modifies the stiffness matrix and optionally the force vector so that K \ f solves the system including boundary conditions.

Everything is showed in the committed cantilevel.ipynb notebook.

codecov-io · 2017-03-23T15:05:34Z

Codecov Report

Merging #125 into master will decrease coverage by 0.9%.
The diff coverage is 84.16%.

@@            Coverage Diff             @@
##           master     #125      +/-   ##
==========================================
- Coverage   88.81%   87.91%   -0.91%     
==========================================
  Files          14       17       +3     
  Lines         948     1274     +326     
==========================================
+ Hits          842     1120     +278     
- Misses        106      154      +48

Impacted Files	Coverage Δ
src/Quadrature/quadrature.jl	`93.33% <ø> (ø)`
src/Quadrature/gaussquad_tri_table.jl	`100% <ø> (ø)`
src/Quadrature/gaussquad_tet_table.jl	`100% <ø> (ø)`
src/deprecations.jl	`0% <ø> (ø)`	⬆️
src/interpolations.jl	`100% <100%> (ø)`	⬆️
src/utils.jl	`100% <100%> (ø)`
src/Export/VTK.jl	`77.5% <100%> (ø)`
src/FEValues/face_values.jl	`100% <100%> (ø)`
src/FEValues/cell_values.jl	`100% <100%> (ø)`
src/Grid/grid.jl	`37.5% <37.5%> (ø)`
... and 16 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 102cbb1...4c1a176. Read the comment docs.

fredrikekre · 2017-04-02T22:04:30Z

src/Dofs/DirichletBoundaryConditions.jl

+# Adds a boundary condition
+function add!(dbcs::DirichletBoundaryConditions, field::Symbol,
+                          nodes::Union{Set{Int}, Vector{Int}}, f::Function, components::Vector{Int})
+    field in dbcs.dh.field_names || error("field $field did not exist in the dof handler, exisitng fields are $(dh.field_names)")


fredrikekre · 2017-04-02T22:12:12Z

src/FEValues/cell_values.jl

 end

-CellVectorValues{dim, FIP <: Interpolation, GIP <: Interpolation}(quad_rule::QuadratureRule{dim}, func_interpol::FIP, geom_interpol::GIP=func_interpol) =
+function CellVectorValues{dim}(quad_rule::QuadratureRule{dim}, func_interpol::Interpolation, geom_interpol::Interpolation=func_interpol)


might as well remove dim parameter here too

fredrikekre · 2017-04-04T07:28:13Z

examples/cantilever.ipynb

+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\u001b[1m\u001b[34mINFO: Recompiling stale cache file /home/kristoffer/.julia/lib/v0.5/JuAFEM.ji for module JuAFEM.\n",


KristofferC · 2017-04-04T08:00:01Z

I dunno why 0.5 is so slow for hyper elasticity...

KristofferC · 2017-04-04T08:03:39Z

Everything is like super slow on 0.5... Stiffness example takes 1.6s on 0.6 and 45s on 0.5...

fredrikekre · 2017-04-04T08:05:50Z

It's probably Ferrite-FEM/Tensors.jl#19

KristofferC · 2017-04-04T08:11:30Z

Ah.. yeah.

KristofferC · 2017-04-04T08:12:16Z

Maybe we should revert it... and just add it to a 0.6 only release later? It makes things pretty unusable on 0.5

fredrikekre · 2017-04-04T08:19:12Z

It's a non-trivial revert at this point. Perhaps just checkout an old commit of Tensors when testing on 0.5?

fredrikekre · 2017-04-04T10:14:12Z

---------------------------------------------------------------------------

                                    Time                   Allocations      

                            ----------------------   -----------------------

      Tot / % measured:          2373s / 99.9%           68.9GiB / 99.9%    

 Section            ncalls     time   %tot     avg     alloc   %tot      avg

 ---------------------------------------------------------------------------

 assemble                6    2354s  99.3%    392s   64.5GiB  93.7%  10.8GiB

   inner assemble    30.0k    2349s  99.0%  78.3ms   64.3GiB  93.4%  2.20MiB

 linear solve            5    17.7s  0.75%   3.54s   4.33GiB  6.28%   886MiB

 export                  1   40.9ms  0.00%  40.9ms   3.49MiB  0.00%  3.49MiB

 ---------------------------------------------------------------------------

such performance 40 min assemble

KristofferC · 2017-04-04T10:17:01Z

I think it is because inlining is disabled on 0.5?

fredrikekre · 2017-04-04T10:18:01Z

How come?

KristofferC · 2017-04-04T10:19:25Z

Because on 0.5 Pkg.test disables inlining because otherwise coverage got bad.

KristofferC · 2017-04-04T10:19:51Z

https://github.com/JuliaLang/julia/blob/release-0.5/base/pkg/entry.jl#L715

fredrikekre · 2017-04-04T10:25:04Z

Oh okay. Perhaps just run the notebooks on 0.6 then, consider them doctests which we also only run on 0.6?

KristofferC · 2017-04-04T10:25:26Z

Yes, that sounds ok!

fredrikekre

merge?

fredrikekre · 2017-04-04T12:02:57Z

Easier to fix small things later than to do more on this branch imo

KristofferC · 2017-04-04T12:05:09Z

Yes, I agree

KristofferC added 15 commits March 21, 2017 13:06

rename boundary to face

8069e65

fix docs

5331920

fixes

c7dcd9c

remove old stuff

d219820

fix name

7ba63e3

stuff

f8d07f2

grid stuff

8f6c6d6

rename boundary to face

285b7c1

add face iterator remove the redundant face list fixup iterators remove stupid getface

move files

d660d48

update for file move

use our names cells for vtk export

33da066

update iterator to hold the current cell and the M parameterization

ce33724

remove some crap

558f2fa

work

e241477

add notebook

7eddaf2

stuff

5d8d59d

KristofferC force-pushed the kc/grid_faces branch from 53e985f to 5d8d59d Compare March 23, 2017 13:44

KristofferC changed the title ~~WIP make faces better~~ Lots of stuff... Mar 23, 2017

KristofferC added 3 commits March 23, 2017 16:01

fix generator for hexahedrons

237a4bb

update with a picture

7c7c0bd

add the pic

3424eb6

KristofferC added 3 commits March 23, 2017 18:56

stuff

f276114

add body load and face load

0443910

add hyperelasticity notebook

1a3b56e

KristofferC force-pushed the kc/grid_faces branch from d5d6d8d to 1a3b56e Compare March 23, 2017 18:56

KristofferC added 2 commits March 23, 2017 20:21

add a heat example

ea1e699

add figures

57aa6f8

fredrikekre reviewed Apr 2, 2017

View reviewed changes

fredrikekre mentioned this pull request Apr 3, 2017

quadratic tetrahedron interpolation #129

Merged

KristofferC added 3 commits April 4, 2017 08:30

adress review and fix which test to disable

8d37f40

reduce number of elements

393bcbf

doc fixes and misc

5926c2e

fredrikekre reviewed Apr 4, 2017

View reviewed changes

fix recomp

402b190

only run notebooks on 0.6

4c1a176

fredrikekre approved these changes Apr 4, 2017

View reviewed changes

KristofferC merged commit 48e39a0 into master Apr 4, 2017

fredrikekre mentioned this pull request Apr 4, 2017

Make Yeoh large deformation notebook work again #91

Closed

kimauth mentioned this pull request Jan 18, 2022

WIP: Traits for ScalarValues / VectorValues #404

Merged

fredrikekre deleted the kc/grid_faces branch January 18, 2022 09:15

fredrikekre mentioned this pull request Apr 11, 2023

Change parameterization of (Abstract)Cell #677

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Lots of stuff... #125

Lots of stuff... #125

KristofferC commented Mar 20, 2017

fredrikekre commented Mar 20, 2017

KristofferC commented Mar 20, 2017

KristofferC commented Mar 23, 2017

codecov-io commented Mar 23, 2017 •

edited

Loading

fredrikekre Apr 2, 2017

fredrikekre Apr 2, 2017

fredrikekre Apr 4, 2017

KristofferC Apr 4, 2017

KristofferC commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre left a comment

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

Lots of stuff... #125

Lots of stuff... #125

Conversation

KristofferC commented Mar 20, 2017

fredrikekre commented Mar 20, 2017

KristofferC commented Mar 20, 2017

KristofferC commented Mar 23, 2017

codecov-io commented Mar 23, 2017 • edited Loading

Codecov Report

fredrikekre Apr 2, 2017

Choose a reason for hiding this comment

fredrikekre Apr 2, 2017

Choose a reason for hiding this comment

fredrikekre Apr 4, 2017

Choose a reason for hiding this comment

KristofferC Apr 4, 2017

Choose a reason for hiding this comment

KristofferC commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

fredrikekre left a comment

Choose a reason for hiding this comment

fredrikekre commented Apr 4, 2017

KristofferC commented Apr 4, 2017

codecov-io commented Mar 23, 2017 •

edited

Loading