MathGroup Archive 2008

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Trig Substitution

  • To: mathgroup at smc.vnet.net
  • Subject: [mg93458] Re: [mg93434] Trig Substitution
  • From: Andrzej Kozlowski <akoz at mimuw.edu.pl>
  • Date: Sun, 9 Nov 2008 05:25:26 -0500 (EST)
  • References: <200811080857.DAA14180@smc.vnet.net> <9517FF96-AEDF-42F5-8F88-84C70505D5EC@mimuw.edu.pl>

On 8 Nov 2008, at 21:45, Andrzej Kozlowski wrote:

>
> On 8 Nov 2008, at 17:57, Bruce Colletti wrote:
>
>> Re Mathematica 6.0.3 under WinXP.
>>
>>
>>
>> Given a sum of (positive integer) powers of sin(t), I wish to cast  
>> (as much
>> as possible) the sum in terms of cos t, chiefly by substituting 1 -  
>> (cos
>> t)^2 for (sin t)^2 throughout.  The code below does this.
>>
>>
>>
>> X=Sin[t]+Sin[t]^2+Sin[t]^3;
>>
>>
>>
>> Y=Apply[List,X]/.{Sin[t]^n_?EvenQ:>Expand[(1-Cos[t]^2)^(n/ 
>> 2)],Sin[t]^n_:>Exp
>> and[(1-Cos[t]^2)^Quotient[n,2] Sin[t]]};
>>
>>
>>
>> Apply[Plus,Y]
>>
>>
>>
>> What is an easier way?  Thankx.
>>
>>
>>
>> Bruce
>>
>
> Obviously this depends on the definition of "easier". If by "easier"  
> you mean a method that is shorter then yes:
>
> gs = GroebnerBasis[{x - (Sin[t] + Sin[t]^2 + Sin[t]^3),
>       Sin[t]^2 + Cos[t]^2 - 1}, {x, Sin[t], Cos[t]}]
> {Cos[t]^2 + Sin[t]^2 - 1, Sin[t]*Cos[t]^2 + Cos[t]^2 + x - 2*Sin[t]  
> - 1}
>
> Solve[Last[gs] == 0, x]
> {{x -> (-Sin[t])*Cos[t]^2 - Cos[t]^2 + 2*Sin[t] + 1}}
>
> If however, by "easier" means that it also should be easy to see why  
> it works, then this method probably does not quality.
>
> Andrzej Kozlowski


Actually, from the point of view of purely "mathematical elegance" it  
might be better to do it without using Solve:

gb = GroebnerBasis[{x - (Sin[t] + Sin[t]^2 + Sin[t]^3),
     Sin[t]^2 + Cos[t]^2 - 1}, {x, Sin[t], Cos[t]}];

Last[PolynomialReduce[x, gb, {x, Sin[t], Cos[t]}]]
(-Sin[t])*Cos[t]^2 - Cos[t]^2 + 2*Sin[t] + 1

But I suspect this does not make really any "easier" in the latter  
sense above.

Andrzej Kozlowski



  • Prev by Date: Re: Trig Substitution
  • Next by Date: Re: Re: bug report: MoebiusMu sum
  • Previous by thread: Re: Trig Substitution
  • Next by thread: Re: Trig Substitution