This is a general problem. The pattern matcher is going to have to be smart enough to recognize things like this, especially for cases that simplify automatically. We should think about how to tackle this.
One idea is to have a set of rules where automatic simplification is known to take place for certain inputs, for instance, (a*b)**x -> a**x*b**x (happens if a and b are numbers, maybe in other cases as well I'm not sure). Then each pattern should run through the rules and generate a new pattern. For instance, (b*log(x))**n would generate b**n*log(x)**n. The rules are then run recursively on this pattern, and so on. I think the sympy.strategies module would be a good fit for this. Then instead of a single pattern you have a list of patterns, and if any match an expression the expression matches. This way, the expression matching itself can be naive and deal only with the structural form of the expression, and be separated from the mathematical rules. That's just one idea. There may be other ways to handle this as well. Aaron Meurer On Thu, Mar 9, 2017 at 12:10 AM, Ankit <[email protected]> wrote: > Hi Aaron ! > > I was thinking in context with the following problem,where I was looking > into the pattern matching. I found this : >>>> pattern = (a+b*sin(c+d*x))**n >>>> term = (1 + 3 * sin(5 + 4 * y))** 10 >>>> pattern.matches(term) > {a_: 1, n_: 10, d_: 4, c_: 5, b_: 3, x_: y} > >>>> pattern = (a*sin(b+c*x))**n >>>> term = (3*sin(5+4*y))**3 >>>> pattern.matches(term) >>>> > > Same goes, with : >>>> pattern2 = (a+ b*log(x))**n >>>> term2 = (3 + 2*log(x))**2 >>>> pattern2.matches(term2) > {a_: 3, n_: 2, b_: 2} > > But, >>>> pattern3 = (b*log(x))**n >>>> term3 = (3*log(x))**2 >>>> > > This is due to the fact,that the term, (3*log(x))**2 is now 9*log(x_)**2. > (Is it a bug? or an improper implementation of pattern matching? If, yes, > how can this be tackled?) > > Due to this, I think special cases have to be considered(while implementing > integration rules) where such case is occurring, during pattern matching. > Please let me know your views on it. > Thanks, > Ankit > > > > On Wednesday, March 8, 2017 at 11:24:21 PM UTC-5, Aaron Meurer wrote: >> >> You can use Pow(10*x, 2, evaluate=False). >> >> Aaron Meurer >> >> On Wed, Mar 8, 2017 at 11:06 PM, Ankit <[email protected]> wrote: >> > Hello ! I have this question ! >> > >> > Python automatically evaluates 10**2 to 100 or say (10*10) to 100 . >> > >> > Same is the case where our expression is , for example: (2*log(x))**2, >> > it >> > will become 4*log(x_)**2 >> > >> > or >> > >> > (10*x)**2 becomes 100*x_**2 (Also, how is x different >> > from >> > x_ ?) >> > >> > Is there anyway,we can stop this evaluation and write these expressions >> > in >> > non-evaluated form. >> > >> > What I thought of was to take them as Strings, and use Evaluate= False, >> > with >> > sympify . But, if this expression is written in Sympy, like expr= >> > (10*x)**2, >> > it will automatically get evaluated and expr = 100*x_**2. >> > >> > Please let me know the possible solution ! >> > >> > Thanks, >> > Ankit >> > >> > -- >> > 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 post to this group, send email to [email protected]. >> > Visit this group at https://groups.google.com/group/sympy. >> > To view this discussion on the web visit >> > >> > https://groups.google.com/d/msgid/sympy/0fe38ea6-3e74-4326-9c78-119b9466e457%40googlegroups.com. >> > For more options, visit https://groups.google.com/d/optout. > > -- > 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 post to this group, send email to [email protected]. > Visit this group at https://groups.google.com/group/sympy. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sympy/cbf027df-b0b8-468d-a2de-97d2bb4099d7%40googlegroups.com. > > For more options, visit https://groups.google.com/d/optout. -- 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 post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAKgW%3D6J9moc1i1LdNsmBY2Hp6Z7LQ3cmbD%3DrzPDA7dvb1%2BqTGA%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
