Master contour integration techniques, understand Cauchy's fundamental theorem, and learn to evaluate complex integrals along curves with path independence.
This follows from Green's theorem. Writing and :
By Green's theorem and the Cauchy-Riemann equations (, ), both integrals vanish.
Problem: Evaluate .
Solution:
Parameterize: for , so
Note: This is non-zero because is not analytic at , which is inside the curve.
Problem: Evaluate where is the straight line from to .
Solution:
Parameterize: for , so
Problem: Verify that using Cauchy's theorem.
Solution:
Since is analytic everywhere (it's an entire function) and the unit disk is simply connected, by Cauchy's theorem, the integral around any closed curve is zero.
Alternatively, parameterizing :
Problem: Evaluate .
Solution:
The function has poles at and , both inside .
By deformation, we can split this into integrals around small circles around each pole, but since the function is not analytic at these points, we'll need Cauchy's integral formula (covered in the next course) or residue theory to evaluate this.
For now, we note that Cauchy's theorem does not apply here because is not analytic at and .
Problem: Evaluate where is any path from to in the right half-plane .
Solution:
In the right half-plane, we can take the principal branch of the logarithm. An antiderivative of is (principal branch).
Therefore:
In simply connected domains, the integral of an analytic function depends only on endpoints, not the path. This is analogous to conservative vector fields in physics.
Cauchy's theorem is the foundation for Cauchy's integral formula, residue theory, and many other powerful results in complex analysis.
The proof using Green's theorem connects complex analysis to real multivariable calculus, showing the deep relationship between these fields.
Problem: Evaluate where is the unit circle traversed once counterclockwise.
Solution:
Parameterize: for , so
Note: This is not zero because has a singularity at , so it's not analytic in a simply connected domain containing .
Problem: Evaluate .
Solution:
Note that . On the unit circle :
when
Since is entire and has residue 1 at (we'll learn this in residue theory), the integral equals .
Problem: Show that if is analytic in a simply connected domain and is a closed curve in , then .
Solution:
If is analytic, then is also analytic (a key result we'll prove later). By Cauchy's theorem, the integral is zero.
Alternatively, if starts and ends at , then by the fundamental theorem of calculus for complex functions.
Problem: Evaluate where is the line segment from to .
Solution:
Parameterize: for , so ,
Note: is not analytic, so the value depends on the path.
Problem: Evaluate along the unit circle .
Solution:
On the unit circle, , so
Parameterize: ,
For circles, use . For line segments, use linear interpolation. For more complex curves, use appropriate parametric equations.
If the function is analytic in a simply connected domain containing the curve, Cauchy's theorem applies and the integral may be zero or path-independent.
If the function has singularities inside the contour, Cauchy's theorem doesn't apply. You'll need residue theory (covered later) or deformation theorem.
Problem: Evaluate where is the circle .
Solution:
Split:
Since is analytic, by Cauchy's theorem.
From Example 4.6,
Therefore, the integral equals .
Problem: Show that where is any simple closed curve around the origin.
Solution:
By deformation theorem, we can shrink to a small circle around the origin. On a small circle, , so the integral approximates .
More precisely, we can use Cauchy's integral formula (covered in the next course) to get the exact result.
Problem: Evaluate .
Solution:
On the unit circle , we have .
The integral is the imaginary part of , which equals 0 (as we'll see with residue theory, or by symmetry).
Problem: If is analytic and on , show that .
Solution:
Using the mean value property (from Cauchy's integral formula),
Taking absolute values:
Problem: Evaluate where is the circle .
Solution:
The function has poles at , both inside .
Using deformation, we can split into integrals around small circles. However, we'll need residue theory (covered later) to evaluate this properly. The result is 0 by the residue theorem since the residues at cancel out.
Problem: Evaluate along any path.
Solution:
Since is entire, the integral is path-independent. An antiderivative is .
Problem: Evaluate .
Solution:
Since is entire, we can use an antiderivative:
Problem: Evaluate .
Solution:
Parameterize: ,
Problem: Show that if is analytic in a domain and , are two paths from to in , and if is simply connected, then .
Solution:
Consider the closed curve (going along then backwards along ).
By Cauchy's theorem:
But
Therefore:
Problem: Evaluate where is the circle .
Solution:
The function has poles at . The circle is centered at with radius 1, so it contains but not .
Using residue theory (to be covered), the integral equals times the residue at .
Problem: Show that if is analytic inside and on a simple closed curve .
Solution:
This is exactly Cauchy's theorem. Since is analytic in a simply connected domain containing and its interior, the integral is zero.
Problem: Show that if is analytic and is a closed curve, then if and only if has no singularities inside .
Solution:
This follows from the residue theorem (covered later). If is analytic inside , then all residues are zero, so the integral is zero.
Problem: Evaluate along the straight line.
Solution:
Since is entire, the integral is path-independent. An antiderivative is .
Problem: Show that where is the circle .
Solution:
On , we have . Using the triangle inequality:
Therefore: on
Problem: Evaluate for integer .
Solution:
Parameterize: ,
This equals if , and 0 otherwise.
Problem: Show that for and any closed curve not containing the origin.
Solution:
If doesn't contain the origin, then is analytic inside (for ) or the origin is outside (for ).
By Cauchy's theorem, if the function is analytic inside , the integral is zero. For , an antiderivative exists, so the integral over a closed curve is zero.
Problem: Parameterize the ellipse and compute along this ellipse.
Solution:
Standard parameterization: for
Then
The integral:
Expanding:
Using and :
Alternatively, since is entire (analytic everywhere), by Cauchy's theorem the integral over any closed curve is zero.
Problem: Parameterize the parabolic arc from to along and compute .
Solution:
Parameterize using , so for
Then , and
The integral:
Expanding
In multiply connected domains, a closed curve might encircle regions where the function is not analytic (holes/singularities). Simply connected domains ensure any closed curve can be shrunk to a point without encountering singularities.
In simply connected domains, Cauchy's theorem implies path independence: since the integral around any closed curve is zero, the integral between two points depends only on the endpoints, not the path taken.
Yes, but only if the closed curve does not enclose any singularities. If a singularity is inside the curve, we need Cauchy's integral formula (covered later) or residue theory.
Express the curve in parametric form $z(t) = x(t) + iy(t)$ for $t$ in some interval, then substitute into the integral and compute $dz = z'(t)dt$. Common parameterizations include circles, line segments, and arcs.
By Cauchy's theorem and deformation principles, you can continuously deform a contour within a region where the function is analytic without changing the integral value, as long as you don't cross any singularities.