Status: Accepted
Owner: asmeurer
CC: smi...@gmail.com
Labels: Type-Defect Priority-High Integration
New issue 2571 by asmeurer: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
This used to work:
In [1]: Integral((exp(x*log(x))*log(x)), x).subs(exp(x*log(x)), x**
Comment #14 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Just a footnote here: xreplace is now available
Integral((exp(x*log(x))*log(x)), x).xreplace({exp(x*log(x)): x**x})
Integral(x**x*log(x), x)
With it, you can shoot
Comment #15 on issue 2571 by asmeu...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
I still think subs should just do this. Substituting terms inside of
integrals is very useful, and this limitation is artificial and unnecessary.
--
You received
Comment #16 on issue 2571 by asmeu...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
That is to say, even for definite integrals it should work. I found myself
wanting to do this quite a bit a while ago when working with solving PDEs
with integral
Comment #17 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Why not just do
i = my_integral
inew = i.func(i.function.subs(old, new), i.limits)
Otherwise you have to come up with an acceptable criteria to govern the
rules of th
Comment #18 on issue 2571 by asmeu...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
I think we already discussed this. That doesn't work so well if the
integral is nested inside an expression.
Let's just do the substitution if it works, and raise
Comment #19 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Regarding the nesting...that's where xreplace can be used where the
transform acts on all Integrals that are in the expression. If you look at
how Ronan implemented
Comment #20 on issue 2571 by asmeu...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
I guess I would expect that to give Integral(y, (y, y**2)).
I get that the user might expect it to do something smart with a change of
variables (e.g., with Integr
Updates:
Labels: NeedsReview smichr
Comment #21 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
https://github.com/sympy/sympy/pull/690
--
You received this message because you are subscribed to the Google Groups
"symp
Updates:
Status: Fixed
Comment #22 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
(No comment was entered for this change.)
--
You received this message because you are subscribed to the Google Groups
"sympy-issues" g
Updates:
Status: Accepted
Labels: -NeedsReview -smichr
Comment #23 on issue 2571 by asmeu...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Sorry, but I still think this isn't powerful enough. The expression from
the original post n
Issue 2571: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
This issue is no longer blocking issue 2010.
See http://code.google.com/p/sympy/issues/detail?id=2010
--
You received this message because you are listed in the owner
or CC fields of this issue, or becaus
Comment #26 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
So could the rule be:
1) allow subs(x, u) always and it changes x to u everywhere
2) allow subs(f(x), f(x)) as long as x does not appear in any limit as a
limit (but
Comment #27 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
I tried to make the first change by inserting as the first line of
Integral._eval_subs, the following:
if old.is_Symbol and new.is_Symbol and not \
Issue 2571: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
This issue is now blocking issue 2010.
See http://code.google.com/p/sympy/issues/detail?id=2010
--
You received this message because you are listed in the owner
or CC fields of this issue, or because you
Comment #2 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Nothing involving dummy symbols is changed anymore. If you look at the
integral with the as_dummy() method you can see what variables are "off
limits":
>>> Int
Comment #3 on issue 2571 by asmeurer: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Well, this needs to be changed. This makes sense for definite integrals,
but not for indefinite integrals, where the integration variable is free.
--
You received this mess
Comment #4 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
I'm willing to look at any sensible rewrite.
--
You received this message because you are subscribed to the Google Groups
"sympy-issues" group.
To post to this group,
Comment #5 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
Should `Integral(x, x).subs(x, 5).doit()` be 25 or 25/2? You get 25 if you
allow subs to target the dummy x of the integrand and the x of the limit
`Integral(x, (x,
Comment #6 on issue 2571 by asmeurer: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
The way I see it, there should be no dummies involved with indefinite
integrals.
On the other hand, it doesn't make sense to allow substitution in the
integration variable
Comment #7 on issue 2571 by renato.c...@gmail.com: Regression in
Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
I think `Integral(x, x).subs(x, 5).doit()` is analogous to
`Derivative(f(x), x).subs(x, 5)`. There's no sense putting the 5 inside the
Integral, so either we
Comment #8 on issue 2571 by asmeurer: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
There's no sense putting the 5 inside the Integral
Yes there is. Why should I be prevented from substituting some nested part
of an expression just because it happens to l
Updates:
Cc: renato.c...@gmail.com
Comment #9 on issue 2571 by asmeurer: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
So I see your issue, though.
There are two ideas of what subs should do. One is evaluating the
expression at a point, and the othe
Comment #10 on issue 2571 by renato.c...@gmail.com: Regression in
Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
There's no sense putting the 5 inside the Integral
Sorry, what I meant here was: there's no sense in integrating wrt a number
(that's why I said it should
Comment #11 on issue 2571 by asmeurer: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
OK, to be clear, I agree that we should not replace
integration/differentiation variables with subs (or anything else where
only a Symbol make sense).
I'm not sure if sub
Comment #12 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
This is what I refer to as 'post-subs semantics', I believe, in the
discussion of what subs should do.
--
You received this message because you are subscribed to the G
Comment #13 on issue 2571 by smi...@gmail.com: Regression in Integral.subs
http://code.google.com/p/sympy/issues/detail?id=2571
And Ronan's comments are probably apropos here: rather than trying to
control this with keywords, perhaps semantically distinc functions
should be created.
--
You rece
27 matches
Mail list logo