On Thu, Jun 17, 2010 at 3:23 PM, Ronan Lamy ronan.l...@gmail.com wrote:
Le jeudi 17 juin 2010 à 14:02 -0700, Ondrej Certik a écrit :
Yes, in this branch we should brake compatibility. We should totally
get rid of the old assumptions, thus breaking compatibility. Make sure
that things work
Le vendredi 18 juin 2010 à 02:02 +0200, Vinzent Steinberg a écrit :
Using the local assumptions hack we could keep compatibility and avoid
rewriting sympy.
I don't see how. Putting everything into global_assumptions works just
as well, it's only less efficient. Changing the assumption system
2010/6/18 Ronan Lamy ronan.l...@gmail.com:
Le vendredi 18 juin 2010 à 02:02 +0200, Vinzent Steinberg a écrit :
Using the local assumptions hack we could keep compatibility and avoid
rewriting sympy.
I don't see how. Putting everything into global_assumptions works just
as well, it's only
Sounds good to me. It should be noted that speeding up the reasoner is
almost entirely disjoint from the assumption interface rewrite --
finished or not, the sat solver improvements don't depend on how you
create your assumptions...just what they are.
Cheers
On Friday, June 18, 2010, Ondrej
On Fri, Jun 18, 2010 at 1:19 PM, Christian Muise
christian.mu...@gmail.com wrote:
Sounds good to me. It should be noted that speeding up the reasoner is
almost entirely disjoint from the assumption interface rewrite --
finished or not, the sat solver improvements don't depend on how you
create
So the main problems start to crop up when you start stripping out
assumptions from the constructor of basic sympy objects. This breaks
compatibility, and very quickly the number of errors goes up. You can view
the natural progression of what you suggest on this branch:
-
I'll see what I can do, but fixing those tests will likely cause more to
fail. I've attached the 7 errors that occur at d25877c:
- Two are because Symbols is trying to use assumptions0, something that was
inherited from AssumeMeths.
- Four are because the tests are trying to create intervals with
Ondrej, if you could provide feedback on these two commits, it would really
help:
-
http://github.com/haz/sympy/commit/91bcb155b24fd354cfdc3b765823ba5041412dce
-
http://github.com/haz/sympy/commit/f0c8becedc87b324f293baea159a73c4057a9a3a
Specifically in the last one, this is something I didn't
2010/6/17 Ondrej Certik ond...@certik.cz:
On Thu, Jun 17, 2010 at 12:27 PM, Christian Muise
christian.mu...@gmail.com wrote:
Ondrej, if you could provide feedback on these two commits, it would really
help:
- http://github.com/haz/sympy/commit/91bcb155b24fd354cfdc3b765823ba5041412dce
-
2010/6/17 Vinzent Steinberg vinzent.steinb...@googlemail.com:
Using global assumptions for is_real will break a lot of thinks,
because global assumptions do not get cleaned automatically. You would
have to clean them after every test setting assumptions.
This assumes you would use global
I think the idea was the compatibility will indeed be broken -- assumptions
are to be stripped from object constructors altogether. At least this is the
impression I was under after speaking with Ondrej.
On Thu, Jun 17, 2010 at 4:43 PM, Vinzent Steinberg
vinzent.steinb...@googlemail.com wrote:
Yes, in this branch we should brake compatibility. We should totally
get rid of the old assumptions, thus breaking compatibility. Make sure
that things work nicely, all tests pass and there are no hacks.
Then, for backwards compatibility, we could introduce the assumptions
into the constructors,
Hi,
I've tried the disconnect-assumptions branch and I got some recursive
exception problems, so I did
git checkout HEAD~2
and then they disappeared, but I still got lots of other exceptions
and failures. Run the test using:
Then this branch here:
http://github.com/certik/sympy/tree/assum
13 matches
Mail list logo