Re: [sympy] nroots and rev-lex

2012-12-31 Thread Aaron Meurer
On Dec 31, 2012, at 3:18 AM, Shriramana Sharma wrote: On Thu, Dec 27, 2012 at 4:37 AM, Aaron Meurer wrote: In the isympy command line I am allowed to set the options of ordering to rev-lex such that I get output of 1 + x + x**2 etc. However it doesn't seem as if isympy supports the ipython Qt

Re: [sympy] nroots and rev-lex

2012-12-31 Thread Shriramana Sharma
On Thu, Dec 27, 2012 at 4:37 AM, Aaron Meurer wrote: >> In the isympy command line I am allowed to set the options of ordering >> to rev-lex such that I get output of 1 + x + x**2 etc. However it >> doesn't seem as if isympy supports the ipython Qt console (please tell >> me if I am mistaken). So

Re: [sympy] nroots and rev-lex

2012-12-29 Thread Aaron Meurer
On Sat, Dec 29, 2012 at 9:31 AM, Shriramana Sharma wrote: > On Sat, Dec 29, 2012 at 1:42 AM, Aaron Meurer wrote: >> The qt console doesn't use real latex yet. Rather it uses matplotlib, which >> actually isn't that great. We plan to include https://github.com/scios/IPyXt >> eventually (patches we

Re: [sympy] nroots and rev-lex

2012-12-29 Thread Shriramana Sharma
On Sat, Dec 29, 2012 at 1:42 AM, Aaron Meurer wrote: > The qt console doesn't use real latex yet. Rather it uses matplotlib, which > actually isn't that great. We plan to include https://github.com/scios/IPyXt > eventually (patches welcome!). Otherwise, just use that directly. OK so I cloned that

Re: [sympy] nroots and rev-lex

2012-12-28 Thread Shriramana Sharma
On Fri, Dec 28, 2012 at 6:48 PM, Shriramana Sharma wrote: > Uh oh, looks like I spake too soon. Durand-Kerner stumbles over itself > in cases of multiple roots being identical. There is already this bug upstream: http://code.google.com/p/mpmath/issues/detail?id=217 -- Shriramana Sharma -- You

Re: [sympy] nroots and rev-lex

2012-12-28 Thread Aaron Meurer
On Dec 28, 2012, at 6:19 AM, Shriramana Sharma wrote: On Fri, Dec 28, 2012 at 3:49 PM, Aaron Meurer wrote: (http://en.wikipedia.org/wiki/Durand-Kerner_method). http://www.cplusplus.com/forum/lounge/82407/. I'm not sure why anyone would want any other method than this, given that it returns a

Re: [sympy] nroots and rev-lex

2012-12-28 Thread Shriramana Sharma
On Fri, Dec 28, 2012 at 3:49 PM, Aaron Meurer wrote: >>> (http://en.wikipedia.org/wiki/Durand-Kerner_method). >> >> http://www.cplusplus.com/forum/lounge/82407/. I'm not sure why anyone >> would want any other method than this, given that it returns all >> real/complex roots quite quickly and with

Re: [sympy] nroots and rev-lex

2012-12-28 Thread Aaron Meurer
On Thu, Dec 27, 2012 at 9:48 PM, Shriramana Sharma wrote: > On Thu, Dec 27, 2012 at 4:37 AM, Aaron Meurer wrote: >> nroots uses mpmath's polyroots, which according to its documentation, >> uses the Durand-Kerner method >> (http://en.wikipedia.org/wiki/Durand-Kerner_method). > > Nice, thanks! I re

Re: [sympy] nroots and rev-lex

2012-12-27 Thread Shriramana Sharma
On Thu, Dec 27, 2012 at 4:37 AM, Aaron Meurer wrote: > nroots uses mpmath's polyroots, which according to its documentation, > uses the Durand-Kerner method > (http://en.wikipedia.org/wiki/Durand-Kerner_method). Nice, thanks! I read up on this and even ended up doing a successful C++ implementati

Re: [sympy] nroots and rev-lex

2012-12-26 Thread Aaron Meurer
On Wed, Dec 26, 2012 at 10:23 AM, Shriramana Sharma wrote: > Two questions mutually unrelated but both related to SymPy. > > What method (of many which seem to be there) does nroots use to > determine the roots of arbitrary degree polynomials? nroots uses mpmath's polyroots, which according to it

[sympy] nroots and rev-lex

2012-12-26 Thread Shriramana Sharma
Two questions mutually unrelated but both related to SymPy. What method (of many which seem to be there) does nroots use to determine the roots of arbitrary degree polynomials? In the isympy command line I am allowed to set the options of ordering to rev-lex such that I get output of 1 + x + x**2