MHB Chebyshev's Method: Recursive Algorithm & Polynomial Form

  • Thread starter Thread starter MarkFL
  • Start date Start date
  • Tags Tags
    Method
AI Thread Summary
Chebyshev's method is a recursive algorithm for calculating the $n$th multiple angle formula for the cosine function, defined as $T_n \equiv \cos(n\theta)$, with the recursion $T_{n+1} = 2xT_n - T_{n-1}$ where $x = \cos(\theta)$. The closed polynomial form for $T_n$ is derived using the characteristic polynomial method, yielding $T_n = \frac{(x + \sqrt{x^2 - 1})^n + (x - \sqrt{x^2 - 1})^n}{2}$. The integral $\int_{-1}^1 T_n \, dx$ evaluates to zero for odd $n$ and to $-\frac{2}{n^2 - 1}$ for even $n$. Additionally, the integral $\int_{-1}^1 \frac{T_n T_m}{\sqrt{1-x^2}} \, dx$ is zero for $n \neq m$ and equals $\frac{\pi}{2}$ for $n = m$.
MarkFL
Gold Member
MHB
Messages
13,284
Reaction score
12
Chebyshev's method is a recursive algorithm for computing the $n$th multiple angle formula for the cosine function. If we define:

$$T_{n}\equiv \cos(n\theta)$$

then the algorithm is given as:

$$T_{n+1}=2xT_{n}-T_{n-1}$$

where:

$$x=\cos(\theta)$$

a) Using trigonometric identities (or otherwise), derive the recursive algorithm.

b) Find a closed polynomial form for $T_{n}$.

c) Compute $$\int_{-1}^1 T_n\,dx$$

d) Compute $$\int_{-1}^1 \frac{T_nT_m}{\sqrt{1-x^2}}\,dx$$
 
Mathematics news on Phys.org
a)

$ \displaystyle T_{n+1}(x) = T_{n+1}(\cos \theta) = \cos \Big((n+1) \theta \Big) = \cos (n \theta) \cos (\theta) - \sin(n \theta) \sin(\theta)$

$ \displaystyle T_{n-1}(x) = T_{n-1}(\cos \theta) = \cos \Big((n-1) \theta \Big) = \cos (n \theta) \cos (\theta) + \sin(n \theta) \sin(\theta) $So $ \displaystyle T_{n+1}(\cos \theta) + T_{n-1} \cos(\theta) = 2 \cos(n \theta) \cos (\theta) = 2 \cos \theta \ T_{n}(\cos \theta) $Or $ \displaystyle T_{n+1}(x) = 2x T_{n}(x) - T_{n-1}(x) $
b)

$ \displaystyle \int_{-1}^{1} T_{n}(x) \ dx $

Let $ x = \cos \theta $

$ \displaystyle = \int_{0}^{\pi} T_{n}(\cos \theta) \sin \theta \ d \theta = \int_{0}^{\pi} \cos(n \theta) \sin \theta \ d \theta = \frac{1}{2} \int_{0}^{\pi} \Big( \sin (n+1) \theta - \sin(n-1) \theta \Big) \ d \theta$

$ \displaystyle = - \frac{1}{2(n+1)} \Big( (-1)^{n+1}- 1 \Big) + \frac{1}{2(n-1)} \Big((-1)^{n-1} -1 \big) $If $n$ is odd, $ \displaystyle \int_{-1}^{1} T_{n}(x) \ dx = 0$

If $n$ is even $ \displaystyle \int_{-1}^{1} T_{n}(x) \ dx = \frac{1}{n+1} - \frac{1}{n-1} = -\frac{2}{n^{2}-1}$
 
Last edited:
d)$ \displaystyle \int_{-1}^1 \frac{T_n(x)T_m(x)}{\sqrt{1-x^2}}\,dx $

Let $x = \cos \theta $

$ \displaystyle = \int_{0}^{\pi} T_n( \cos \theta)T_m(\cos \theta ) \ d \theta = \int_{0}^{\pi} \cos( n \theta) \cos( m \theta) \ d \theta = \frac{1}{2} \int_{0}^{\pi} \Big(\cos(n+m) \theta + \cos (n-m) \Big) \ d \theta $If $n \ne m$, $ \displaystyle \int_{-1}^1 \frac{T_n(x)T_m(x)}{\sqrt{1-x^2}}\,dx = \frac{1}{2(n+m)} \Big(\sin(n+m) \pi - 0 \Big) + \frac{1}{2(n-m)} \Big( \sin(n-m) \pi - 0 \Big) = 0 $If $n = m$, $ \displaystyle \int_{-1}^1 \frac{T^{2}_n(x)}{\sqrt{1-x^2}}\,dx = \frac{1}{4n} \Big(\sin(2n \pi) - 0 \Big) + \frac{\pi}{2} = \frac{\pi}{2} $
 
Last edited:
I skipped (b) because I initially thought we were dealing with a recurrence relation with a variable coefficient. (Doh)$ \displaystyle T_{n+1} - 2xT_{n} - T_{n-1} = 0 $ has the associated characteristic polynomial $r^{2}-2xr - 1 = 0$

which has roots $ \displaystyle r = x \pm \sqrt{x^{2}-1} $.So the general solution is $T_{n} = c_{1} \Big(x + \sqrt{x^{2}-1} \Big)^{n} + c_{2} \Big(x - \sqrt{x^{2}-1} \Big)^{n} $.$ \displaystyle T_{0} = 1 = c_{1}+c_{2}$

$\displaystyle T_{1} = x = c_{1} \Big(x + \sqrt{x^{2}-1} \Big) + c_{2} \Big(x - \sqrt{x^{2}-1} \Big) $So just by inspection, $\displaystyle c_{1}=c_{2} = \frac{1}{2} $

And $ \displaystyle T_{n} = \frac{\Big(x + \sqrt{x^{2}-1} \Big)^{n} + \Big(x - \sqrt{x^{2}-1} \Big)^{n}}{2} $ which is evidently a polynomial in $x$ for all values of $n$
 
Last edited:
Bravo, Random Variable! (Clapping)

Thank you for taking the time to post such lucid explanations. (Yes)

Here are my solutions:

a) Using trigonometric identities (or otherwise), derive the recursive algorithm.

We may begin with the identity:

$$\cos((n+1)\theta)=\cos((n+1)\theta)$$

Add $$0=\cos((n-1)\theta)-\cos((n-1)\theta)$$ to the right side:

$$\cos((n+1)\theta)=\cos((n+1)\theta)+\cos((n-1)\theta)-\cos((n-1)\theta)$$

To the first two terms on the right, apply the following sum to product identity:

$$\cos(\alpha)+\cos(\beta)=2\cos\left(\frac{\alpha-\beta}{2} \right)\cos\left(\frac{\alpha+\beta}{2} \right)$$

and we have:

$$\cos((n+1)\theta)=2\cos(\theta)\cos(n\theta)-\cos((n-1)\theta)$$

And so we may write:

$$T_{n+1}=2xT_{n}-T_{n-1}$$

b) Find a closed polynomial form for $T_{n}$.

This is a homogeneous recurrence whose associated auxiliary equation is:

$$r^2-2xr+1=0$$

Application of the quadratic formula yields:

$$r=x\pm\sqrt{x^2-1}$$

Thus, the solution is of the form:

$$T_n(x)=c_1\left(x-\sqrt{x^2-1} \right)^n+c_2\left(x+\sqrt{x^2-1} \right)^n$$

Use initial conditions to determine constants:

$$T_0=c_1+c_2=1$$

$$T_1=c_1\left(x-\sqrt{x^2-1} \right)+c_2\left(x+\sqrt{x^2-1} \right)=x$$

Substituting from the first into the second:

$$c_1\left(x-\sqrt{x^2-1} \right)+\left(1-c_1 \right)\left(x+\sqrt{x^2-1} \right)=x$$

$$c_1x-c_1\sqrt{x^2-1}+x+\sqrt{x^2-1}-c_1x-c_1\sqrt{x^2-1}=x$$

$$-2c_1\sqrt{x^2-1}+\sqrt{x^2-1}=0$$

$$c_1=\frac{1}{2}\,\therefore\,c_2=\frac{1}{2}$$

Thus, the closed form for $T_n(x)$ is:

$$T_n(x)=\frac{1}{2}\left(\left(x-\sqrt{x^2-1} \right)^n+\left(x+\sqrt{x^2-1} \right)^n \right)$$

Using the binomial theorem, we may state:

$$T_n(x)=\frac{1}{2}\left(\sum_{k=0}^n\left({n \choose k}x^{n-k}\left(-\sqrt{x^2-1} \right)^k \right)+\sum_{k=0}^n\left({n \choose k}x^{n-k}\left(\sqrt{x^2-1} \right)^k \right) \right)$$

$$T_n(x)=\frac{1}{2}\left(\sum_{k=0}^n\left({n \choose k}x^{n-k}(-1)^k\left(\sqrt{x^2-1} \right)^k \right)+\sum_{k=0}^n\left({n \choose k}x^{n-k}\left(\sqrt{x^2-1} \right)^k \right) \right)$$

$$T_n(x)=\frac{1}{2}\sum_{k=0}^n\left(\left(1+(-1)^k \right){n \choose k}x^{n-k}\left(\sqrt{x^2-1} \right)^k \right)$$

$$T_n(x)=\sum_{k=0}^{\left\lfloor \frac{n}{2} \right\rfloor}\left({n \choose 2k}x^{n-2k}\left(x^2-1 \right)^k \right)$$

c) Compute $$\int_{-1}^1 T_n\,dx$$

From the recurrence relation, we may compute:

$$T_n(x)=\frac{1}{2}\left(\frac{1}{n+1}\cdot\frac{d}{dx}T_{n+1}(x)-\frac{1}{n-1}\cdot\frac{d}{dx}T_{n-1}(x) \right)$$

$$\int_{-1}^1 T_n(x)\,dx=\frac{1}{2}\left[\frac{T_{n+1}(x)}{n+1}-\frac{T_{n-1}(x)}{n-1} \right]_{-1}^1$$

From the definition, we see that $T_n(1)=1$ and $T_n(-1)=1\text{ n even or }-1\text{ n odd}$, thus:

For even $n$:

$$\int_{-1}^1 T_n(x)\,dx=\frac{1}{n+1}-\frac{1}{n-1}=\frac{2}{1-n^2}$$

For odd $n$:

$$\int_{-1}^1 T_n(x)\,dx=0$$

d) Compute $$\int_{-1}^1 \frac{T_nT_m}{\sqrt{1-x^2}}\,dx$$

$$Let x=\cos(\theta)\,\therefore\,dx=-\sin(\theta)\,d\theta$$

$$I=\int_{0}^{\pi} \cos(n\theta)\cos(m\theta)\,d\theta$$

For $n=m=0$:

$$I=\int_{0}^{\pi}\,d\theta=\pi$$

For $n=m\ne0$:

$$I=\int_{0}^{\pi} \cos^2(n\theta)\,d\theta$$

Let $$u=n\theta\,\therefore\,du=n\,d\theta$$

$$I=\frac{1}{n}\int_{0}^{n\pi} \cos^2(u)\,du=\frac{1}{2n}\int_{0}^{n\pi}\cos(2u)+1\,du=$$

$$\frac{1}{2n}\left[\frac{1}{2}\sin(2u)+u \right]_0^{n\pi}=\frac{1}{2n}\left(0+n\pi-0-0 \right)=\frac{\pi}{2}$$

For $n\ne m$:

$$I=\int_{0}^{\pi} \cos(n\theta)\cos(m\theta)\,d\theta=$$

$$\left[\frac{m\sin(m\theta)\cos(n\theta)-n\cos(m\theta)\sin(n\theta)}{m^2-n^2} \right]_{0}^{\pi}=$$

$$\frac{1}{m^n-n^2}(0-0-0+0)=0$$
 
Insights auto threads is broken atm, so I'm manually creating these for new Insight articles. In Dirac’s Principles of Quantum Mechanics published in 1930 he introduced a “convenient notation” he referred to as a “delta function” which he treated as a continuum analog to the discrete Kronecker delta. The Kronecker delta is simply the indexed components of the identity operator in matrix algebra Source: https://www.physicsforums.com/insights/what-exactly-is-diracs-delta-function/ by...
Fermat's Last Theorem has long been one of the most famous mathematical problems, and is now one of the most famous theorems. It simply states that the equation $$ a^n+b^n=c^n $$ has no solutions with positive integers if ##n>2.## It was named after Pierre de Fermat (1607-1665). The problem itself stems from the book Arithmetica by Diophantus of Alexandria. It gained popularity because Fermat noted in his copy "Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos, et...
Thread 'Imaginary Pythagorus'
I posted this in the Lame Math thread, but it's got me thinking. Is there any validity to this? Or is it really just a mathematical trick? Naively, I see that i2 + plus 12 does equal zero2. But does this have a meaning? I know one can treat the imaginary number line as just another axis like the reals, but does that mean this does represent a triangle in the complex plane with a hypotenuse of length zero? Ibix offered a rendering of the diagram using what I assume is matrix* notation...

Similar threads

Back
Top