1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Complex function holomorphic

  1. Feb 19, 2017 #1
    1. The problem statement, all variables and given/known data
    Suppose z = x + iy. Where are the following functions differentiable? Where are they holomorphic? Which are entire?

    the function is f(z) = e-xe-iy
    2. Relevant equations
    ∂u/∂x = ∂v/∂y

    ∂u/∂y = -∂v/∂x

    3. The attempt at a solution
    f(z) = e-xe-iy

    I convert it to polar form:

    f(z) = cos(xy) + isin(xy)

    then i set u as the real part of the equation, and v as the complex part of the equation.

    u = cos(xy)
    v = isin(xy)

    ∂u/∂x = -ysin(xy)
    ∂v/∂y= ixcos(xy)

    then ∂u/∂x =/= ∂v/∂y so the function is not holomorphic.

    I think i did this wrong because i've done this for the majority of the problem set, and every function is coming up as not holomorphic. Is my work correct?

    EDIT: I did it again and here is my new work, my other work was wrong because i multiplied exponents wrong

    f(z) = e-xe-iy

    = (cos(-x) + isin(-x)) * (cos(-y) + isin(-y))
    = cos(-x)cos(-y) + icos(-y)sin(-x) + (-1)sin(-x)(sin-y) + isin(-x)cos(-y)

    so Re(f(z)) = cos(-x)cos(-y) - sin(-x)sin(-y) = u
    Im(f(z)) = cos(-y)sin(-x) + cos(-x)sin(-y) = v

    then ux = -(-sin(-x)cos(-y)) + cos(-x)sin(-y)

    ux = sin(-x)cos(-y) + cos(-x)sin(-y)

    and vy = - (-sin(-y)sin(-x)) + cos(-x)cos(-y)

    vy = sin(-y)sin(-x) - cos(-x)cos(-y)

    so ux =/= vy

    so f(z) is not holomorphic, and is not differentiable at all points( but may be differentiable at some points ), and is not entire because it is not differentiable at all points.
    Last edited: Feb 19, 2017
  2. jcsd
  3. Feb 19, 2017 #2


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    It is holomorphic. There must be an error in your calculation. The calculations might be simpler if you leave e-x as it is, since it is a real function whose derivative you know.
  4. Feb 20, 2017 #3
    I just want to confirm though, that if a complex function satisfies the Cauchy Riemann equations, then the complex function is also holomorphic. So my logic is right, but I probably just differentiated incorrectly somewhere?
  5. Feb 20, 2017 #4


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    With one caveat. The first partial derivatives of u and v must be continuous.
  6. Feb 20, 2017 #5
    Thank you for the help. I think my error is when I'm converting from exponential to polar form.

    f(z) = e-xe-iy

    = e-iy-x

    and now I need to rewrite -iy-x so that i can pull out the i, and have a real angle for my polar form.

    So I know z = x+iy, then I can say -z = -iy - x

    so I can rewrite: f(z) = e-z

    i'm unsure of whether I can treat z as a single variable and then say that e raised to any single variable will be differentiable at all points.

    Am I on the right path here? I feel like i've made a couple assumptions in differentiating e-z where z ∈ ℂ.
  7. Feb 20, 2017 #6


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I don't think you need to make that big a change. It takes you straight into complex analysis and complex derivatives.

    Just to make the calculations of your original approach easier, try:
    f(z) = e-xe-iy
    = e-x×(cos(-y) + isin(-y))
    = e-x × cos(-y) + i*e-x ×sin(-y)
    The partial derivatives of the real and imaginary parts are probably easier in that form.
  8. Feb 20, 2017 #7


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Have you already established in class that ez is an entire function?
    If so, the easiest proof is:
    1) g(z) = -z and h(z)=ez are both entire functions.
    2) The composition of entire functions is entire. So f(z) = h(g(z)) is entire.

    If those facts have not already been established, your original approach is the right thing to do.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted