Bug in translation to LaTeX: double supperscript error

[ma...@mendelu.cz]

Hello, this topis has been started on september on sage-support (no
answer yet). I think that sage-devel is more suitable for this
problem. I repost it here with some comments (workaround and minimal
example for the bug)

On 4 zář, 12:11, "ma...@mendelu.cz" <ma...@mendelu.cz> wrote:
> Hello, this code give "doubleexponent error".
>
> var('a x b gamma p')
> f(x)=a*x^((gamma-p))+b*x^((-p))
> (f((b*p/((gamma-p)*a))^(1/gamma))).show()
>

Since another user reported similar behavior to me, I am sending more
comments.

workaround is to simplify like this
maxima(f((b*p/((gamma-p)*a))^(1/gamma))).radcan().sage().show()

minimal example for the problem is
sage: latex((x^pi)^e)
{(x)}^{\pi}^{e} --- this is not valid TeX expression and jsmath
does not render it.


I tried to fix it ad it seems that this has to be fixed in
/libs/ginac/decl.pxi
somewhere close to
object GEx_to_str_latex "_to_PyString_latex<ex>"(GEx *s) except +
but i do not uderstand the format and the purpouse of this file.

Perhaps somebody else has better idea and more experiences with this.

Robert


- Hide quoted text -

> Is it error in sage to TeX converter?
> Tested today on sagenb.org
>
> Thanks
> Robert Marik

[Burcin Erocal]

Hi Robert,

On Fri, 6 Nov 2009 02:24:53 -0800 (PST)
"ma...@mendelu.cz" <ma...@mendelu.cz> wrote:

>
> Hello, this topis has been started on september on sage-support (no
> answer yet). I think that sage-devel is more suitable for this
> problem. I repost it here with some comments (workaround and minimal
> example for the bug)
>
> On 4 zář, 12:11, "ma...@mendelu.cz" <ma...@mendelu.cz> wrote:
> > Hello, this code give "doubleexponent error".
> >
> > var('a x b gamma p')
> > f(x)=a*x^((gamma-p))+b*x^((-p))
> > (f((b*p/((gamma-p)*a))^(1/gamma))).show()
> >
>
> Since another user reported similar behavior to me, I am sending more
> comments.
>
> workaround is to simplify like this
> maxima(f((b*p/((gamma-p)*a))^(1/gamma))).radcan().sage().show()
>
> minimal example for the problem is
> sage: latex((x^pi)^e)
> {(x)}^{\pi}^{e} --- this is not valid TeX expression and jsmath
> does not render it.
>
>
> I tried to fix it ad it seems that this has to be fixed in
> /libs/ginac/decl.pxi
> somewhere close to
> object GEx_to_str_latex "_to_PyString_latex<ex>"(GEx *s) except +
> but i do not uderstand the format and the purpouse of this file.
>
> Perhaps somebody else has better idea and more experiences with this.

Thanks for taking the time to isolate the issue. The code for printing
symbolic expressions is in pynac (C++). The fix can be as simple as
printing an extra set of braces around power objects.

If anybody wants to try fixing this, the relevant function is
power::do_print_latex() in power.cpp. To get to the file (using the
instructions I wrote in another message just now), go to your SAGE_ROOT
and do:

./sage -f -s spkg/standard/pynac-0.1.9.p0.spkg

cd spkg/build/pynac-0.1.9/src/ginac

Edit power.cpp. To compile and make your changes effective, go to your
SAGE_ROOT again, and do

./sage -sh
cd spkg/build/pynac-0.1.9/src
make install

Of course, another option is to open a ticket for this if there isn't
one already and I'll take a look at it when I find some time.

Thanks.

Burcin

[ma...@mendelu.cz]

Hello Burcin, thanks for detailed explanation where to fix this issue.

> "ma...@mendelu.cz" <ma...@mendelu.cz> wrote:
>
> Thanks for taking the time to isolate the issue. The code for printing
> symbolic expressions is in pynac (C++). The fix can be as simple as
> printing an extra set of braces around power objects.
>

I think that this is not sifficient: we have to clearly distinguish
between (a^b)^c and a^(b^c). Hence we need not only braces, but also
parentheses in some cases. Compare the following three LaTeX strings

${a^b}^c$

$a^{b^c}$

$\left({a^b}\right)^c$


You can use either http://www.tlhiv.org/ltxpreview/ to see that there
is only minor difference btween the first and the second output.

I guess that we need to fix also power::print_power but do not have
experisnces in C.

I will at leas check if there exists a trac abvout this and if not, I
open this issue on trac server.

Many thanks
Robert

[ma...@mendelu.cz]

Do no know how to test if the base of power is another power and if it
is sufficient to ad braces, or if also parentheses should be added.
Hence I only opened the ticket at http://trac.sagemath.org/sage_trac/ticket/7406

Thank you for the hints.

Robert

On 6 lis, 13:00, "ma...@mendelu.cz" <ma...@mendelu.cz> wrote:
> I think that this is not sufficient: we have to clearly distinguish