A fixed point iteration method for the arctangent with any odd order of convergence based on sine and cosine

2026

Read A fixed point iteration method for the arctangent with any odd order of convergence based on sine and cosine (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-06698.txt.

Introduction

Alois Schiessl presents a fixed point iteration for computing the arctangent, the inverse tangent function, built from sine and cosine evaluations. The iteration subtracts a finite sum of odd powers of a sine-cosine ratio from the current estimate. Starting close enough to the true value, the sequence converges to the arctangent with an order that the user can choose in advance by setting the number of sum terms.

Specifically, using P terms gives convergence order exactly two P plus one, so accuracy accelerates sharply with each added term. A computational test reported in the paper demonstrates the method's efficiency in practice. The work is a compact numerical recipe with a tunable speed guarantee.

Catalog record

YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.06698

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A fixed point iteration method for the arctangent with any odd order of convergence based on sine and cosine Abstract In this paper, we present a fixed point method for the arctangent based on sine and cosine. Let t ∈ ℝ + t\in\mathbb{R}^{+} and P ∈ ℕ P\in\mathbb{N} . We define: T ⁡ ( x ) = x − ∑ k = 1 P ( − 1 ) k − 1 2 ​ k − 1 ​ ( sin ⁡ ( x ) − t ​ cos ⁡ ( x ) cos ⁡ ( x ) + t ​ sin ⁡ ( x ) ) 2 ​ k − 1 . T\left(x\right)=x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2\,k-1}. For every initial value x 0 x_{0} sufficiently close to arctan ⁡ ( t ) \arctan\left(t\right) , the sequence x n + 1 = T ( x n ) ; n = 0,1 , … x_{n+1}=T\left(x_{n}\right)\;;\,n=0,1,\ldots is converging to arctan ⁡ ( t ) \arctan\left(t\right) with order of convergence exactly ( 2 ​ P + 1 ) \left(2\,P+1\right) . The computational test we performed demonstrates the efficiency of the method. Zusammenfassung In dieser Abhandlung stellen wir ein Fixpunktverfahren zur Berechnung des arcustangens auf Basis von sinus und cosinus vor. Es sei t ∈ ℝ + t\in\mathbb{R}^{+} und P ∈ ℕ P\in\mathbb{N} . Wir definieren: T ⁡ ( x ) = x − ∑ k = 1 P ( − 1 ) k − 1 2 ​ k − 1 ​ ( sin ⁡ ( x ) − t ​ cos ⁡ ( x ) cos ⁡ ( x ) + t ​ sin ⁡ ( x ) ) 2 ​ k − 1 . T\left(x\right)=x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2\,k-1}. Für jeden Startwert x 0 x_{0} hinreichend nahe bei arctan ⁡ ( t ) \arctan\left(t\right) konvergiert die Folge x n + 1 = T ( x n ) ; n = 0,1 , … x_{n+1}=T\left(x_{n}\right)\;;\,n=0,1,\ldots gegen arctan ⁡ ( t ) \arctan\left(t\right) mit Konvergenzordnung genau ( 2 ​ P + 1 ) \left(2\,P+1\right) . Anhand einer praktischen Berechnung von π 4 \frac{\pi}{4} zeigen wir die Effizienz des Verfahrens. Deutsche Version ab Seite 17 Alois Schiessl 1 1 1 University of Regensburg E-Mail: [email protected] August 10, 2026 1 Introduction and main result We start with the \arctan power series [ 1 ] \arctan\left(u\right)=u-\frac{u^{3}}{3}+\frac{u^{5}}{5}-\frac{u^{7}}{7}+-\dots=\sum_{k=1}^{\infty}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}\,u^{\,2\,k-1}\,. The power series is absolutely convergent for \left|u\right|<1 . We only use the first P terms: u-\frac{u^{3}}{3}+\frac{u^{5}}{5}-\frac{u^{7}}{7}+-\ldots+\frac{\left(-1\right)^{P-1}}{2\,P-1}\,u^{2\,P-1}=\sum_{k=1}^{P}\frac{\left(-1\right)^{k-1}}{2\,k-1}\,u^{2\,k-1}\,. Let t\in\mathbb{R}^{+} and substituting u=\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)} leads to the finite sum \sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2\,k-1}. With this sum we define a fixed point function that enables the calculation of \arctan\left(t\right) with any odd order of convergence. Since \arctan\left(-t\right) = -\arctan\left(t\right) , it is sufficient considering only positiv values. Now let’s formulate our statement: Theorem . Let t\in{\mathbb{R^{+}}} and P\in{\mathbb{N}} . We define the function T:\;\;[\,0\,,\frac{\pi}{2}]\;\;\rightarrow\mathbb{R}\;; \quad x\mapsto T\left(x\right)\;; T\left(x\right)=x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2\,k-1}. Then, the following statements hold, for all t\in\mathbb{R}^{+}: (a) T\left(x\right) is a fixed point function with \arctan\left(t\right) as fixed point; that means T\bigl(\arctan\left(t\right)\bigr)=\arctan\left(t\right)\,. (b) For the derivatives at the fixed point applies : \displaystyle T^{\left(k\right)}{\bigl(\arctan\left(t\right)\bigr)}=\left\{\begin{aligned} 0\quad\qquad\qquad\qquad\quad 0\leq&\,k\leq 2\,P\\ \\ \left(-1\right)^{P}\left(2\cdot P\right)\,!\qquad\qquad\quad k=\,&2\,P+1\\ \end{aligned}\right. (c) For every initial value x_{0} sufficiently close to \arctan\left(t\right) , the sequence \displaystyle x_{n+1} \displaystyle=T\left(x_{n}\right)\,,\;n=0,1,\,\ldots is converging to \arctan\left(t\right) with order of convergence exactly \left(2\,P+1\right)\,. 2 Proof of the Theorem In this section, we prove the statements of the theorem. In order to simplify the notation for T\left(x\right)=x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2\,k-1}, we define the function: Let t\in\mathbb{R^{+}} , f:\;\;[0,\frac{\pi}{2}]\;\;\rightarrow\mathbb{R}\;; f\left(x\right)=\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\,. The fixed point function is then easier to write as T\left(x\right)=x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}f\left(x\right)^{2\,k-1} 2.1 Proof of fixed point We must prove that the following statement is true at the fixed point \arctan\left(t\right) : T\bigl(\arctan\left(t\right)\bigr)=\arctan\left(t\right). It suffices to prove that the following relation holds f\bigl(\arctan\left(t\right)\bigr)=\frac{\sin\!\bigl(\arctan\left(t\right)\bigr)-t\cos\!\bigl(\arctan\left(t\right)\bigr)}{\cos\!\bigl(\arctan\left(t\right)\bigr)+t\sin\!\bigl(\arctan\left(t\right)\bigr)}=0\,. From a mathematical handbook [ 4 ] , we obtain the following statements : \sin\,\bigl(\arctan\left(t\right)\bigr)=\frac{t}{\sqrt{1+t^{2}}}\;\;;\;t\in\mathbb{R}\;; \cos\,\bigl(\arctan\left(t\right)\bigr)=\frac{1}{\sqrt{1+t^{2}}}\;\;;\;t\in\mathbb{R}\;; Plugging in gives f\bigl(\arctan\left(t\right)\bigr)=\frac{\frac{t}{\sqrt{1+t^{2}}}-t\frac{1}{\sqrt{1+t^{2}}}}{\frac{1}{\sqrt{1+t^{2}}}+t\frac{t}{\sqrt{1+t^{2}}}}=\frac{\frac{t}{\sqrt{1+t^{2}}}-\frac{t}{\sqrt{1+t^{2}}}}{\frac{1}{\sqrt{1+t^{2}}}+\frac{t^{2}}{\sqrt{1+t^{2}}}}=\frac{0}{\frac{1}{\sqrt{1+t^{2}}}+\frac{t^{2}}{\sqrt{1+t^{2}}}}=0\,. From f\bigl(\arctan\left(t\right)\bigr)=0 follows directly that T\bigl(\arctan\left(t\right)\bigr)=\arctan\left(t\right)-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}{\underbrace{f\bigl(\arctan\left(t\right)\bigr)}_{=0}}^{2\,k-1}=\arctan\left(t\right)\,. This proves the fixed point property. 2.2 Proof of the derivatives Next, we turn to the derivatives. This requires a little bit more work. 2.2.1 First Derivative Let’s start with the first derivative. We use the fixed point function in the representation T\left(x\right)=x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}f\left(x\right)^{2\,k-1} with f\left(x\right)=\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\,. Differentiating the fixed point function yields \displaystyle T\,^{\prime}\left(x\right) \displaystyle=\left(x-\sum_{k=1}^{P}\,\frac{\left(-1\right)^{k-1}}{2\,k-1}f\left(x\right)^{2\,k-1}\right)^{\prime} \displaystyle=1-\sum_{k=1}^{P}\left(-1\right)^{k-1}f\left(x\right)^{2\,k-2}f\,^{\prime}\left(x\right)\,. \displaystyle=1+\sum_{k=1}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2}f\,^{\prime}\left(x\right)\,. We need the derivative of f\left(x\right)=\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\,. We leave this task to the computer algebra system Maple and trust in the result: f\,^{\prime}\left(x\right)=1+\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2}=1+f\left(x\right)^{2}. Substituting this into the first derivative, we get \displaystyle T\,^{\prime}\left(x\right) \displaystyle=1+\sum_{k=1}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2}\left(1+f\left(x\right)^{2}\right) \displaystyle=1+\sum_{k=1}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2}+\sum_{k=1}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k}. To make the matter more transparent, we will refer to the first sum as \displaystyle s_{1}=1+\sum_{k=1}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2} and the second sum as s_{2}=\sum_{k=1}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k} The first derivative can then be written simply as T\,^{\prime}\left(x\right)=s_{1}+s_{2}\,. Next step is splitting the sum s_{1} as follows: \displaystyle s_{1} \displaystyle=1+\sum_{k=1}^{1}\left(-1\right)^{k}f\left(x\right)^{2\,k-2}+\sum_{k=2}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2} \displaystyle=1-1+\sum_{k=2}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2} \displaystyle=\sum_{k=2}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k-2}. Now shifting the summation index leads to \displaystyle s_{1} \displaystyle=\sum_{k=1}^{P-1}\left(-1\right)^{k+1}f\left(x\right)^{2\,k} \displaystyle= \displaystyle-\sum_{k=1}^{P-1}\left(-1\right)^{k}f\left(x\right)^{2\,k}\,. Next, let’s look at s_{2} . Also splitting the sum and simplifying gives a more simplified form: \displaystyle s_{2} \displaystyle=\sum_{k=1}^{P-1}\left(-1\right)^{k}f\left(x\right)^{2\,k}+\sum_{k=P}^{P}\left(-1\right)^{k}f\left(x\right)^{2\,k} \displaystyle=\sum_{k=1}^{P-1}\left(-1\right)^{k}f\left(x\right)^{2\,k}+\left(-1\right)^{P}f\left(x\right)^{2\,P}. Up to now we have \displaystyle s_{1} \displaystyle=-\sum_{k=1}^{P-1}\left(-1\right)^{k}f\left(x\right)^{2\,k} \displaystyle s_{2} \displaystyle=\;\;\;\sum_{k=1}^{P-1}\left(-1\right)^{k}\bigl(f\left(x\right)\bigr)^{2\,k}+\left(-1\right)^{P}f\left(x\right)^{2\,P}\,. When used, we get \displaystyle T\,^{\prime}\left(x\right)= \displaystyle\;\;\;s_{1}+s_{2} \displaystyle= \displaystyle-\sum_{k=1}^{P-1}\left(-1\right)^{k}f\left(x\right)^{2\,k} \displaystyle+\sum_{k=1}^{P-1}\left(-1\right)^{k}f\left(x\right)^{2\,k}+\left(-1\right)^{P}f\left(x\right)^{2\,P} \displaystyle=\left(-1\right)^{P}f\left(x\right)^{2\,P}\,. The sums are vanishing and it remains the simple term: T\,^{\prime}\left(x\right)=\left(-1\right)^{P}f\left(x\right)^{2\,P}=\left(-1\right)^{P}\left(\frac{\sin\!\left(x\right)-t\cos\!\left(x\right)}{\cos\!\left(x\right)+t\sin\!\left(x\right)}\right)^{2\,P}\,. Now plugging in the fixed point \arctan\left(t\right) gives us T\,^{\prime}\bigl(\arctan\left(t\right)\bigr)=\left(-1\right)^{P}\left(\frac{\sin\bigl(\arctan\left(t\right)\bigr)-t\cos\bigl(\arctan\left(t\right)\bigr)}{\cos\bigl(\arctan\left(t\right)\bigr)+t\sin\bigl(\arctan\left(t\right)\bigr)}\right)^{2}\,. We have already calculated the term in brackets in the fixed point proof: \frac{\sin\bigl(\arctan\left(t\right)\bigr)-t\cos\bigl(\arctan\left(t\right)\bigr)}{\cos\bigl(\arctan\left(t\right)\bigr)+t\sin\bigl(\arctan\left(t\right)\bigr)}=0\,. Plugging in, finally results in T\,^{\prime}\bigl(\arctan\left(t\right)\bigr)=0\,. Up to now, we have the particular results at the fixed point T\,\bigl(\arctan\left(t\right)\bigr)=\arctan\left(t\right)\,; T\,^{\prime}\bigl(\arctan\left(t\right)\bigr)=0\,. 2.2.2 Higher derivatives We still have to prove that the following statement holds at the fixed point \arctan\left(t\right): \displaystyle T^{\left(k\right)}{\bigl(\arctan\left(t\right)\bigr)}=\left\{\begin{aligned} 0\;\quad\qquad\qquad\;\;1\leq&\,k\leq 2\,P\\ \\ \left(-1\right)^{P}\left(2\,P\right)\,!\qquad\quad k=&2\,P+1\\ \end{aligned}\right. We must differentiate \left(2\,P+1\right) times T\left(x\right) and then plugging in \arctan\left(t\right) . Using the first derivative T\,^{\prime}\left(x\right) , the task simplifies to \displaystyle\left.T^{\left(2\,P+1\right)}\left(x\right)\right|_{x=\arctan\left(t\right)}=\left.\left(T\,^{\prime}\right)^{\left(2\,P\right)}\left(x\right)\right|_{x=\arctan\left(t\right)}=\left.\left(\left(-1\right)^{P}f\left(x\right)^{2\,P}\right)^{\left(2\,P\right)}\right|_{x=\arctan\left(t\right)}\,. On the right-hand side, we need at \arctan\left(t\right) the \left(2\,P\right)^{th} derivative of f\left(x\right)^{2\,P} . Remembering f\bigl(\arctan\left(t\right)\bigr)=0 and calculating f^{\prime}\bigl(\arctan\left(t\right)\bigr)=1+f\left(\arctan\left(t\right)\right)^{2}=1\,. Then, we proceed as follows: We expand f\left(x\right) into a Taylor series of degree 1 at the fixed point \arctan\left(t\right) with Lagrange remainder: f\left(x\right)=f\bigl(\arctan\left(t\right)\bigr)+f^{\prime}\bigl(\arctan\left(t\right)\bigr)\bigl(x-\arctan\left(t\right)\bigr)+\frac{f\,^{\prime\prime}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)^{2}\,. where \xi is between x and \arctan\left(t\right) . Since, in accordance with f\bigl(\arctan\left(t\right)\bigr)=0 and f^{\prime}\bigl(\arctan\left(t\right)\bigr)=1 , we obtain f\left(x\right)=\bigl(x-\arctan\left(t\right)\bigr)+\frac{f\,^{\prime\prime}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)^{2}\,. Factoring out \bigl(x-\arctan\left(t\right)\bigr) yields f\left(x\right)=\bigl(x-\arctan\left(t\right)\bigr)\left(1+\frac{f\,^{\prime\prime}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)\,. Next raising the equation to the \left(2\,P\right)^{\,th} power: f\left(x\right)^{2\,P}=\bigl(x-\arctan\left(t\right)\bigr)^{2\,P}\left(1+\frac{f{\,{}^{\prime\prime}}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)^{2\,P}\,. We calculate the higher derivatives of f\left(x\right)^{2\,P} by using the Leibniz rule [ 4 ] . Leibniz rule states that if u\left(x\right) and v\left(x\right) are n -times differentiable functions, then the product u\left(x\right)\cdot v\left(x\right) is also n -times differentiable and its n^{\,th} derivative is given by the sum \bigl(u\left(x\right)\cdot v\left(x\right)\bigr)^{\left(n\right)}=\sum\limits_{k=0}^{n}\binom{n}{k}\,u\left(x\right)^{\left(k\right)}\cdot v\left(x\right)^{\left(n-k\right)}\,. Here we want to understand u\left(x\right)^{\left(0\right)}=u\left(x\right) and v\left(x\right)^{\left(0\right)}=v\left(x\right)\,. In our case we have: n=2\,P\;; u\left(x\right)=\bigl(x-\arctan\left(t\right)\bigr)^{2\,P}\;; v\left(x\right)=\left(1+\frac{f\,^{\prime\prime}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)^{2\,P}\;; Thus we obtain f\left(x\right)^{2\,P}=\underbrace{{\bigl(x-\arctan\left(t\right)\bigr)}^{2\,P}}_{u\left(x\right)}\cdot\underbrace{{{\left(1+\frac{f\,^{\prime\prime}{\left(\xi\right)}}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)}^{2\,P}}}_{v\left(x\right)}\,. Now applying Leibniz rule gives \displaystyle\left(f\left(x\right)^{2\,P}\right)^{\left({2\,P}\right)} \displaystyle=\sum\limits_{k=0}^{{2\,P}}\binom{{2\,P}}{k}\,\left(\bigl(x-\arctan\left(t\right)\bigr)^{2\,P}\right)^{\left(k\right)} \displaystyle\times\left(\left(1+\frac{f\,^{\prime\prime}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)^{{2\,P}}\right)^{\left({2\,P-k}\right)} We first focus on the sequence of derivatives u\left(x\right)^{\left(k\right)}=\left(\bigl(x-\arctan\left(t\right)\bigr)^{{2\,P}}\right)^{\left\{k\right\}},k=0,1,\ldots,{2\,P}\,. From a mathematical handbook [ 4 ] , we derive the following formula: \displaystyle\Bigl(\bigl(x-\arctan\left(t\right)\bigr)^{2\,P}\Bigr)^{\left(k\right)}=\frac{\left(2\,P\right)\,!}{\left(2\,P-k\right)\,!}\bigl(x-\arctan\left(t\right)\bigr)^{2\,P-k}\,,k=0,1,\ldots,{2\,P}. In particular, we note that for k<2\,P in u\left(x\right)^{\left(k\right)}=\left(\bigl(x-\arctan\left(t\right)\bigr)^{{2\,P}}\right)^{\left(k\right)} the factor \bigl(x-\arctan\left(t\right)\bigr) always occurs and thus for x=\arctan\left(t\right) the terms are vanishing. Only the case k=\left(2\,P\right) leads to a non-zero result, namely the product {\left.u{{\left(x\right)}^{\left(2\,P\right)}}\right|}_{x=\arctan\left(t\right)}=\left(2\,P\right)\,!\,. From the sum \sum\limits_{k=0}^{{2\,P}}\binom{{2\,P}}{k}\,\left(\bigl(x-\arctan\left(t\right)\bigr)^{2\,P}\right)^{\left(k\right)}\cdot\left(\left(1+\frac{f\,^{\prime\prime}{\left(\xi\right)}}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)^{{2\,P}}\right)^{\left({2\,P-k}\right)}, it only rests the term with k=\left(2\,P\right) at the fixed point \arctan\left(t\right) . We can evaluate \displaystyle{\left.{{\left(f{{\left(x\right)}^{2\,P}}\right)}^{\left(2\,P\right)}}\right|}_{x=\arctan\left(t\right)} \displaystyle=\underbrace{\sum\limits_{k={2\,P}}^{{2\,P}}\binom{{2\,P}}{{2\,P}}\left.\left(\bigl(x-\arctan\left(t\right)\bigr)^{2\,P}\right)^{\left(2\,P\right)}\right|_{x=\arctan\left(t\right)}}_{=\left(2\,P\right)\,!} \displaystyle\times\underbrace{\left.\left(\left(1+\frac{f\,^{\prime\prime}\left(\xi\right)}{2}\bigl(x-\arctan\left(t\right)\bigr)\right)^{2\,P}\right)\right|_{x=\arctan\left(t\right)}}_{=1}\,. We obtain the simple result {\left.{{\left(f{{\left(x\right)}^{2\,P}}\right)}^{\left(2\,P\right)}}\right|}_{x=\arctan\left(t\right)}=\left(2\,P\right)\,!\,. Our task was the calculation of \displaystyle\left.T\left(x\right)^{\left(2\,P+1\right)}\right|_{x=\arctan\left(t\right)}=\left.T\,^{\prime}\left(x\right)^{\left(2\,P\right)}\right|_{x=\arctan\left(t\right)}=\left.\left(\left(-1\right)^{P}f\left(x\right)^{2\,P}\right)^{\left(2\,P\right)}\right|_{x=\arctan\left(t\right)}\,. Now plugging in yields T\bigl(\arctan\left(t\right)\bigr)^{\left(2\,P+1\right)}=\left(-1\right)^{P}\left(2\,P\right)\,! This completes the proof of the derivations. 2.3 Proof of convergence We do the prove by using Banach’s fixed point theorem

Read the complete text