Many thanks for the remarks. Together, I see the possibility to improve the
readability (and maybe the performance) of my code in two ways:
1) If I can get rid of the exec statements, the code will be much easier to
read.
2) By using two symbols for each variable, I can improve the readability
and maybe the performance. Specifically, I want to define one symbol that
is exactly the same on the LHS and the RHS of symbols, then a second symbol
where the RHS is a LaTeX format.
For example
Sicj means S suffix i comma j. Sicj will be used in expressions and SicjL
will be used (only) for printing.
Here are some code attempts:
for ind in range(1,maxIter+1):
for ind2 in range(1,maxIter+1):
strSym = f'S{ind}c{ind2} = symbols(\'S{ind}c{ind2}\')'
exec(strSym, globals())
#strSym = f'S{ind}c{ind2}L = symbols(\'S_{{ind}\\,{ind2}}\')'
strSym = 'S%dc%dL = symbols(\'S_{%d\\,%d}\')' %
(ind,ind2,ind,ind2)
exec(strSym, globals())
Remarks:
I didn't get any of the suggestions working for two indices. It looks like
the colon in symbols always introduces one number which is always at the
end of the expression.
As a result, I didn't find a way to eliminate the exec statements for
symbols with two indices.
strSym = f'S{ind}c{ind2}L = symbols(\'S_{{ind}\\,{ind2}}\')' leads to an
error, and I didn't find a way to use { and } to refer to Python string
manipulation and LaTeX string manipulation without getting a conflict.
On Thursday, February 11, 2021 at 8:30:30 PM UTC+1 [email protected] wrote:
> Another option here is numbered_symbols(), which returns a generator
> which produces any number of numbered symbols. It's also useful in
> conjunction with the take() function:
>
> >>> gen = numbered_symbols('a')
> >>> n = 10
> >>> take(gen, n)
> [a0, a1, a2, a3, a4, a5, a6, a7, a8, a9]
>
> In general, if you want to create symbols in complicated ways that
> doesn't match the convenience methods of symbols(), you may be better
> of generating them manually using the Symbol constructor and a list
> comprehension or an explicit for loop:
>
> >>> [Symbol("a%d" % i) for i in range(n)]
> [a0, a1, a2, a3, a4, a5, a6, a7, a8, a9]
>
> Aaron Meurer
>
> On Thu, Feb 11, 2021 at 5:22 AM Oscar Benjamin
> <[email protected]> wrote:
> >
> > On Thu, 11 Feb 2021 at 11:19, Thomas Ligon <[email protected]> wrote:
> > >
> > > "The solution to this in Python is to use lists or tuples or some
> other container rather than raw variables. For example:
> > >
> > > x = symbols('x:10')"
> > >
> > > Based on that, and
> > > Core — SymPy 1.7.1 documentation
> > >
> > > a = symbols('a:2*maxIter+1')
> > >
> > > The result is
> > >
> > > (a0*maxIter+1, a1*maxIter+1)
> > >
> > > The problem is that symbols does not support variables after the
> colon, defeating the whole purpose of using indexed symbols to begin with.
> >
> > You can use the features of the Python language to generate the string
> > as input to the symbols function:
> >
> > >>> num_symbols = 10
> >
> > This is using %-formatting of strings:
> >
> > >>> 'x:%d' % num_symbols
> > 'x:10'
> > >>> symbols('x:%d' % num_symbols)
> > (x0, x1, x2, x3, x4, x5, x6, x7, x8, x9)
> >
> > This is using f-strings:
> >
> > >>> f'x:{num_symbols}'
> > 'x:10'
> > >>> symbols(f'x:{num_symbols}')
> > (x0, x1, x2, x3, x4, x5, x6, x7, x8, x9)
> >
> > > I am trying to use indexed symbols.
> >
> > Note that "indexed symbols" means something different in sympy e.g.:
> >
> > >>> from sympy import IndexedBase
> > >>> x = IndexedBase('x')
> > >>> x
> > x
> > >>> x[0]
> > x[0]
> >
> > IndexedBase represents a mathematical quantity that can be subscripted
> > but in a symbolic way. For example if I have some data points that I
> > refer to as x_1, x_2, etc. and I want to refer to x_i then I can use
> > IndexedBase and index it with a symbol.
> >
> > Oscar
> >
> > --
> > You received this message because you are subscribed to the Google
> Groups "sympy" group.
> > To unsubscribe from this group and stop receiving emails from it, send
> an email to [email protected].
> > To view this discussion on the web visit
> https://groups.google.com/d/msgid/sympy/CAHVvXxRCL%3D2cQKoGTJjUos6onXEWAujitgDwZaY3Ohhn2ONmvA%40mail.gmail.com
> .
>
--
You received this message because you are subscribed to the Google Groups
"sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/sympy/4469b4d6-d3a7-4478-b932-8a865efd6715n%40googlegroups.com.
