Closed dpsanders closed 7 years ago
Merging #220 into master will decrease coverage by
1.54%
. The diff coverage is40.47%
.
@@ Coverage Diff @@
## master #220 +/- ##
==========================================
- Coverage 91.33% 89.78% -1.55%
==========================================
Files 26 26
Lines 1050 1067 +17
==========================================
- Hits 959 958 -1
- Misses 91 109 +18
Impacted Files | Coverage Δ | |
---|---|---|
src/intervals/functions.jl | 95.74% <ø> (ø) |
:white_check_mark: |
src/intervals/arithmetic.jl | 100% <ø> (+1.32%) |
:white_check_mark: |
src/ValidatedNumerics.jl | 100% <ø> (ø) |
:white_check_mark: |
src/intervals/rounding.jl | 41.02% <39.47%> (-35.45%) |
:x: |
src/intervals/precision.jl | 84.61% <50%> (-5.04%) |
:x: |
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 de9e9d2...980e350. Read the comment docs.
I changed RoundingType
--> IntervalRounding
for clarity and precision
I think the CRlibm problem was because it requires CRlibm v0.5. This should now be fixed.
This branch is aimed only at post-Julia-0.4.
This is ready. All tests pass on 0.5, and all tests except multidimensional pass on 0.6; these are waiting for StaticArrays improvements.
I added a basic test in interval_tests/rounding.jl for changing the rounding mode. This checks that the sin function is correctly redefined on 0.6 (and is only run on 0.6) and tests pass there. The tests are tricky to get right due to "world-age" problems (the fix of Julia issue #265 on redefining dependent functions.)
I would like to merge and release 0.8.
Merging...
This is an implementation of changing the global rounding mode using "traits".
Note that actually changing the rounding mode for calculations works only on Julia v0.6, since it depends on the solution to https://github.com/JuliaLang/julia/issues/265.
From the point of view of the user, it reduces to doing
I have not yet added tests.