I think it should stay the way it is now. Functions should not convert floats to rationals automatically in general, unless they absolutely must. "In the face of ambiguity, refuse the temptation to guess." and also "Explicit is better than implicit."
If you want it to call nsimplify() only when is_rational is used, this is impossible, as the sqrt is already evaluated at that point. Aaron Meurer On Sat, Mar 3, 2012 at 9:59 PM, prateek papriwal <papriwalprat...@gmail.com> wrote: > i found that >>>>sqrt(4.41) > 2.10000000 > > i guess we have no problem with this . > > the problem we have is with .is_rational (rational checking function) > > so when sqrt(4.41).is_rational is called , we should call > > sqrt(nsimplify(4.41,rational=True)).is_rational > > which gives True in sympy ... > > also sqrt(4.41).is_rational would be more user friendly term than > sqrt(nsimplify(4.41,rational=True)).is_rational > > On Sun, Mar 4, 2012 at 10:24 AM, prateek papriwal > <papriwalprat...@gmail.com> wrote: >> >> so somehow we need to merge nsimplify() method into sqrt() method . >> >> in that case when we call sqrt(4.41) >> >> automatically sqrt(nsimplify(4.41,rational=True)) will be called . >> >> >> >> >> On Sun, Mar 4, 2012 at 10:07 AM, Aaron Meurer <asmeu...@gmail.com> wrote: >>> >>> Exactly. This happens automatically for rational numbers, so that if >>> the number is a perfect square, then they will be reduced (indeed, it >>> will reduce any perfect square factors if the number is not too big). >>> sqrt() of a Float will just give you another Float, so if you want to >>> test that, you have to convert it to a Rational first. Normally, I >>> would suggest calling Rational on it to do that, but that won't work >>> because of issue http://code.google.com/p/sympy/issues/detail?id=2950. >>> >>> So instead, you should use nsimplify(rational=True), like >>> >>> In [58]: nsimplify(4.41, rational=True) >>> Out[58]: >>> 441 >>> ─── >>> 100 >>> >>> In [59]: sqrt(nsimplify(4.41, rational=True)) >>> Out[59]: >>> 21 >>> ── >>> 10 >>> >>> Aaron Meurer >>> >>> On Sat, Mar 3, 2012 at 9:09 PM, prateek papriwal >>> <papriwalprat...@gmail.com> wrote: >>> > we can do the following thing (talking about square roots) >>> > >>> > For integer inputs, only the square roots of the square numbers are >>> > rationals. So our problem boils down to find if our number is a square >>> > number . in this way sqrt(3) can be checked . >>> > If we have rational numbers as inputs (that is, a number given as the >>> > ratio >>> > between two integer numbers), check that both divisor and dividend are >>> > perfect squares. in this way 4.41 can be checked . >>> > >>> > Finally we know that any finite floating point number is a rational >>> > number >>> > .. >>> > On Sun, Mar 4, 2012 at 5:05 AM, Aaron Meurer <asmeu...@gmail.com> >>> > wrote: >>> >> >>> >> On Sat, Mar 3, 2012 at 12:17 PM, Sergiu Ivanov >>> >> <unlimitedscol...@gmail.com> wrote: >>> >> > On Sat, Mar 3, 2012 at 6:29 PM, prateek papriwal >>> >> > <papriwalprat...@gmail.com> wrote: >>> >> >> yes u were right i had an old version 6.7 , now i have a new >>> >> >> version >>> >> >> which >>> >> >> gives "NONE" for >>> >> >> >>> >> >>>>>sqrt(3).is_rational >>> >> >> and >>> >> >> >>> >> >>>>>sqrt(3).is_irrational >>> >> >> >>> >> >> this need to be corrected >>> >> > >>> >> > Yes, this would be highly desired, but, as Aaron has said, there is >>> >> > no >>> >> > simple general way to check the rationality of an expression. >>> >> > >>> >> > Could you please describe what you are trying to achieve? If you >>> >> > can >>> >> > narrow down your problem sufficiently well, you may be able to >>> >> > devise >>> >> > an ad-hoc way to check the rationality of an expression in your >>> >> > domain. >>> >> > >>> >> > Sergiu >>> >> >>> >> Problems only arise if you have transcendental numbers, or if you have >>> >> symbolic expressions with some assumptions on them. If you are >>> >> dealing with a non-symbolic algebraic number, it is always possible to >>> >> tell if it's rational or not. One way is to use minpoly() and see if >>> >> the minimal polynomial is linear or not. We don't currently do this >>> >> because minpoly() is too slow for non-trivial algebraic numbers (if I >>> >> remember correctly). >>> >> >>> >> Aaron Meurer >>> >> >>> >> -- >>> >> You received this message because you are subscribed to the Google >>> >> Groups >>> >> "sympy" group. >>> >> To post to this group, send email to sympy@googlegroups.com. >>> >> To unsubscribe from this group, send email to >>> >> sympy+unsubscr...@googlegroups.com. >>> >> For more options, visit this group at >>> >> http://groups.google.com/group/sympy?hl=en. >>> >> >>> > >>> > -- >>> > You received this message because you are subscribed to the Google >>> > Groups >>> > "sympy" group. >>> > To post to this group, send email to sympy@googlegroups.com. >>> > To unsubscribe from this group, send email to >>> > sympy+unsubscr...@googlegroups.com. >>> > For more options, visit this group at >>> > http://groups.google.com/group/sympy?hl=en. >>> >>> -- >>> You received this message because you are subscribed to the Google Groups >>> "sympy" group. >>> To post to this group, send email to sympy@googlegroups.com. >>> To unsubscribe from this group, send email to >>> sympy+unsubscr...@googlegroups.com. >>> For more options, visit this group at >>> http://groups.google.com/group/sympy?hl=en. >>> >> > > -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To post to this group, send email to sympy@googlegroups.com. > To unsubscribe from this group, send email to > sympy+unsubscr...@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/sympy?hl=en. -- You received this message because you are subscribed to the Google Groups "sympy" group. To post to this group, send email to sympy@googlegroups.com. To unsubscribe from this group, send email to sympy+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sympy?hl=en.
