search

JuliaApproximation / CompactBases.jl

Julia library for function approximation with compact basis functions
MIT License
16 stars 2 forks source link

Overhaul of finite-differences #41

Closed jagot closed 3 years ago

codecov[bot] commented 3 years ago

Codecov Report

Merging #41 (3bdcdec) into master (f184728) will decrease coverage by 4.78%. The diff coverage is 47.60%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master      #41      +/-   ##
==========================================
- Coverage   69.77%   64.99%   -4.79%     
==========================================
  Files          21       22       +1     
  Lines        1360     1551     +191     
==========================================
+ Hits          949     1008      +59     
- Misses        411      543     +132
Impacted Files Coverage Δ
src/CompactBases.jl 88.88% <ø> (ø)
src/bsplines.jl 76.08% <0.00%> (-0.56%) :arrow_down:
src/fedvr.jl 87.03% <0.00%> (-0.41%) :arrow_down:
src/restricted_bases.jl 76.74% <0.00%> (+2.93%) :arrow_up:
src/coulomb_derivative.jl 40.00% <40.00%> (ø)
src/fd_derivatives.jl 48.28% <43.11%> (-20.37%) :arrow_down:
src/finite_differences.jl 70.31% <65.15%> (-8.02%) :arrow_down:
src/materialize_dsl.jl 86.79% <0.00%> (-0.49%) :arrow_down:
... and 2 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 f184728...3bdcdec. Read the comment docs.

jagot commented 3 years ago

@dlfivefifty Julia 1.5 dies on Windows, for some reason (in addition to the non-fatal https://github.com/JuliaLang/julia/issues/37532 present on all platforms in 1.5, but fixed in 1.6). Do we care?

dlfivefifty commented 3 years ago

Can you try any @test_broken to see if there's a StackOverflow not being picked up?

jagot commented 3 years ago

Hmm, the only test marked as broken is this: https://github.com/JuliaApproximation/CompactBases.jl/blob/b5d36286f63f0484140d45a149e8576a6144873c/test/bsplines/splines.jl#L104

In any case, the crash occurs before any of the tests have even run: https://github.com/JuliaApproximation/CompactBases.jl/runs/2191648905?check_suite_focus=true#step:6:123

I wonder if this is related to the memory spike that @mortenpi sees in #38?

signal (22): SIGABRT
in expression starting at none:0
terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc
dlfivefifty commented 3 years ago

Yes it could be an out-of-memory issue, the Windows machine might not be as flexible. Perhaps just remove the Windows tests until someone complains?

jagot commented 3 years ago

Closes #40 Closes #42