0ce89254e0
update docs and call this 0.19
2014-06-10 16:27:36 +02:00
909c19dc24
update url
2014-04-29 00:55:41 +02:00
26c98b05a1
update urls
2014-04-29 00:55:12 +02:00
e2a1e65eee
update url
2014-04-29 00:53:39 +02:00
c621cc84c4
Merge pull request #289 from skirpichev/sympy-asserts
...
Don't use asserts in the code
2014-04-25 22:00:30 +02:00
2b0eb67a22
Don't use asserts in the code (partial fix from sympy/sympy#5090 )
2014-04-25 22:11:28 +04:00
e88aef80ad
Merge pull request #288 from skirpichev/move-issues
...
Move issues from the Google Code
2014-04-24 13:59:52 +02:00
ab3e2cacf9
Move issues: shift issue numbers (1 -> 40)
2014-04-20 22:08:51 +04:00
e43bc402bc
Move issues: change issue tracker URI
2014-04-20 21:24:30 +04:00
2fcac5cbe0
credit skirpichev
2014-04-16 17:38:39 +02:00
8d91dc94b2
Merge pull request #21 from skirpichev/travis-doctests
...
WIP: Enable doctests in Travis-CI
2014-04-16 16:24:49 +02:00
861a76896a
identify(): Use sorted(), while working with dictionary arguments (fix doctest)
2014-04-16 17:58:20 +04:00
164ddd793f
SKIP rounding "doctests"
2014-04-16 17:58:20 +04:00
d29966edb8
Fix doctests for matrices
2014-04-16 17:58:20 +04:00
062e62de6c
SKIP sympy doctests
2014-04-16 17:58:20 +04:00
3f01fc326b
py3 fixes for doctests
2014-04-16 17:58:20 +04:00
ef942890f8
Fix tests in linalg.py
2014-04-16 17:58:19 +04:00
6711eed731
Use ctx.zero in polyroots()
2014-04-16 17:58:19 +04:00
eecbb0e63d
Fix mpi_to_str doctest
2014-04-16 17:58:19 +04:00
d8f074e245
Workaround for pypy in the steffensen's doctest
2014-04-16 17:58:19 +04:00
006c2c843f
Fix steffensen doctest
2014-04-16 17:50:38 +04:00
a162d13f0c
Change imports for optimization module
2014-04-16 17:50:38 +04:00
0983147d83
disable py.test colors for travis
2014-04-16 17:50:38 +04:00
71ec9537da
Enable doctests in Travis-CI
2014-04-15 15:06:30 +04:00
5e27d1dda5
convert plot code to a proper script
2014-04-15 12:52:33 +02:00
35379e92ba
Update a couple of URLs in README.rst
2014-04-15 12:25:11 +02:00
e6aaecc42c
Merge pull request #20 from certik/line-endings
...
Fix line endings
2014-04-15 10:35:35 +02:00
89c636f5dc
Normalize line endings
2014-04-14 12:49:26 -06:00
971afb65ba
Fix line endings handling in git
...
See [1] for more information.
[1] https://help.github.com/articles/dealing-with-line-endings
2014-04-14 12:48:20 -06:00
9207fb2f7f
Merge pull request #19 from certik/docs
...
polyroots: clarify a docstring about extraprec
2014-04-13 13:43:49 +02:00
5fb8631b3a
docs: add a missing word
2014-04-12 14:24:26 -06:00
ce98e63970
polyroots: clarify a docstring about extraprec
2014-04-10 16:17:05 -06:00
0f02c47336
Merge pull request #15 from certik/polyroots
...
Make polyroots more robust
2014-04-10 22:01:22 +02:00
d42f331fc1
polyroots: Improve documentation
...
We did few changes here and there and added a new section about how the
precision works.
2014-04-10 13:45:53 -06:00
23cf69cf77
Use workdps context manager in the test
2014-04-10 12:28:23 -06:00
d39d6fa52c
Use extraprec context manager
...
Instead of try/finally blocks.
2014-04-10 12:25:34 -06:00
cbe4181609
Use +ctx.eps instead of ctx.eps*1
2014-04-10 11:17:11 -06:00
970d548f49
Add a test for NoConvergence
2014-04-10 11:00:43 -06:00
3166859c15
Add a test for polyroots
2014-04-10 10:56:21 -06:00
fdb3284a42
Merge pull request #18 from skirpichev/travis-banner
...
Add travis banner in the README.rst
2014-04-10 16:57:11 +02:00
3abbc7e484
Add travis banner in the README.rst
2014-04-10 18:49:20 +04:00
6e3e5a3436
Merge pull request #17 from skirpichev/enable-travis
...
Enable travis-ci builds
2014-04-10 16:40:14 +02:00
b8d2f3422f
Enable Travis-CI tests, closes #16
2014-04-10 17:34:56 +04:00
18a9ddb308
Don't use py.magic to figure out rootdir
2014-04-10 17:34:54 +04:00
a2e7c3aeff
make polyroots more robust
...
* fix tolerance calculation
We now set the tolerance `tol` directly equal to `ctx.eps`. This is the right
way to do it, since if more working precision is needed, the user needs to
assign the `extraprec` argument.
* use `extraprec` argument
Previously this argument was not used at all by a mistake and instead the
default value 10 was assigned directly.
* chop small numbers in "cleanup" mode
Previously only small real or imaginary parts were chopped. But if abs(x) is
close to zero, we want to chop this case as well.
* Raise NoConvergence error if convergence is not achieved
* always refine all roots
Previously if the given root was already converged, it wasn't refined. But this
seems to increase the number of iterations needed to converge considerably.
2014-04-09 15:33:19 -06:00
4ce0b0d1db
call this 0.18
2013-12-31 19:41:53 +01:00
cf24b399a4
update changelog
2013-12-28 17:30:43 +01:00
4cf4110ef5
fix fourier() documentation error
2013-12-28 16:53:09 +01:00
e46e492df9
update a number
2013-12-26 17:41:55 +01:00
5c7ecd171b
update a date
2013-12-26 17:39:37 +01:00
Fredrik Johansson