

Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Solutions to various math problems, including best polynomial approximation and complex integrals. It covers topics such as the maclaurin series, maximum modulus theorem, residues, and cauchy's integral formula. Students studying advanced mathematics, particularly complex analysis, will find this document useful.
Typology: Assignments
1 / 3
This page cannot be seen from the preview
Don't miss anything!


0. Exhibit a function ƒ(z) , continuous (but not analytic) throughout the closed unit disk, to which function ƒ(z) no polynomial can approximate arbitrarily closely thereon.
Answer 0 : Try ƒ(z) := 1 – |z|^2. It is continuous. But no polynomial P can satisfy |ƒ(z) – P(z)| < ß < 1/4 throughout the unit disk. This is so because P would have to satisfy ß > |ƒ(z) – P(z)| = |P(z)| on the disk’s boundary |z| = 1 , and then the Maximum Modulus Theorem would imply ß > |P(0)| too, forcing |ƒ(0) – P(0)| = |1 – P(0)| > 1 – ß > ß.
Polynomials can easily approximate arbitrarily closely to any function ƒ(z) analytic throughout the closed unit disk. Then the Maclaurin series for ƒ converges in a larger disk of radius bigger than some r > 1 , and converges in the closed unit disk at least as fast as a geometric series with common ratio 1/r < 1; this means that the series’ first N terms constitute a polynomial that approximates ƒ as closely as desired if N is chosen big enough.
The situation is more interesting when ƒ is analytic in the open unit disk but merely continuous towards and along its boundary, or when ƒ is analytic throughout a closed bounded simply connected region with a piecewise smooth boundary. But those are stories for another day about best polynomial approximation.
1(a). As z traverses the unit circle O once counterclockwise, how many times does z^2 – 1/
go around the origin? What is ∫ O 2z·dz/(z^2 – 1/4)? Answers : Twice. 4π ı.
1(b). As Φ increases from 0 to 2π , the expression ( e^3 ı Φ^ – 1/2)(2 e –2 ı Φ^ + 1) traces some closed curve in the complex plane. How many times does that curve go around the origin?
Answer : Once because (z^3 – 1/2)(2/z^2 + 1) has 3 zeros and 2 poles inside the unit circle.
2. Evaluate ∫E z^7 · e sin(z)^ ·dz integrated around the ellipse E traced by z = 5·cos(t) + 2 ı ·sin(t)
as t increases from 0 to 2π. Answer : 0 , because the integrand is analytic inside and on E.
what contour you use and which singularities you take into account. Answers : π and π/. The contour is an infinite semicircle above its diameter, the real axis; the singularity is a pole at z = ı ; the residue there is – ı /2 for the first and – ı /4 for the second integral because
The next six problems are intended to be solved in sequence.
5(a). Check that cot(z) = cos(z)/sin(z) has a simple pole at z = nπ for every integer n , and prove that cot(z) has no other singularities in the finite complex z-plane. What is the residue of cot(z) at each of its poles? Answer : When 1/cot(z) = tan(z) = 0 its derivative
1 + tan^2 (z) = 1 ≠ 0 so the zeros of tan are all simple poles of cot. Those zeros z = nπ are the
same as the zeros of 2 ı · e ı z^ sin(z) = e 2ı z^ – 1 all of which must be real because otherwise
6(a). For every integer n ≥ 0 let Cn be the boundary, traversed counterclockwise, of the z-
plane’s square in which neither |Re(z)| nor |Im(z)| exceeds nπ + π/2. Calculate ∫C n cot(z)·dz.
Answer : ∫C n cot(z)·dz = 2π ı (2n+1) = 2π ı ∑ ( Residues of cot(z) inside Cn ).
7. Find a constant upper bound M for |cot(z)| on all of Cn by first treating each of its sides
separately. Of course, there are many choices available for M ; the smaller the better. Use it to show for any fixed complex w that ∫C n cot(z)dz/((z–w)z) → 0 as n → +∞. Answer : On
|∫C n cot(z)dz/((z–w)z)| < M·(Cn ’s Perimeter)/((nπ+π/2 – |w|)(nπ+π/2)) = 8M/(nπ+π/2 – |w|) → 0.
8. What is the residue of cot(z)/(z – w) at each of its poles? Recall problem 5 to see how the answer depends upon whether w is a pole of cot. Suppose now that w lies inside Cn ; what
is ∫C n cot(z)dz/(z–w) and why is it not 2π ı ·cot(w)? Answer : If w ≠ nπ the residue of
cot(z)/(z–w) at z = nπ is 1/(nπ – w) , and the residue at z = w is cot(w) ; but if w = nπ = z the residue at z is 0. If w lies inside Cn , Cauchy’s Integral Formula says that 2π ı ·ƒ(w) =
∫C n ƒ(z)dz/(z–w) when ƒ is analytic on and inside Cn , as cot is not. Instead, provided w/π
is not an integer, ∫C n cot(z)dz/(z–w) = 2π ı ·( cot(w) + ∑–n≤k≤n 1/(kπ – w) ). This is an analytic
function of w with removable singularities at w = mπ for integers m between ±n inclusive, for which ∫C n cot(z)dz/(z–mπ) = 2 ı ·∑–n≤k≤n & k≠m 1/(k – m).
that this series for cot(z) converges uniformly in any compact region that contains no pole of cot. What does this imply for term-by-term integration of the series? ( It’s no power series.)
= ∫C n cot(w)dw/(w–z) by Problem 6b …