# Teaching a new class this semester

Starting next week I'll be teaching a new class (Introduction to Dynamical Oceanography). I'll be using the Introduction to Geophysical Fluid Dynamics, 2nd Edition book by Cushman-Roisin & Beckers. This book has an interesting approach where all the topics are shown from an analytical and numerical point of view.

However, there is one big issue with the numerical exercises, all the numerical experiments examples are MatlabTM code.

Well, since I have to read and learn this examples in order to prepare my lectures, so why not convert them to python while I do it?

Here is the first program from chapter 1 aliasanim.m, a MatlabTM function that shows an animation of the aliasing problem. It samples a cosine function with increasing time steps.

In [2]:
import numpy as np
import matplotlib.pyplot as plt
from matplotlib import animation
from JSAnimation import IPython_display

def aliasanim(n=300, nper=2):
"""Function that shows an animation of the aliasing problem.
It samples a cosine function with increasing time steps.
n: number of initial sampling points per period of the signal
nper: number of periods shown for the real signal (suggested: n=300, nper=2)
when launched, the sampling frequency is decreased."""

twopi = 2 * np.pi
k = np.arange(0, nper * n)
# Initial signal, nicely sampled if n is large.
# di = omega Delta t
di = twopi / (nper*n) * nper
t = di * k
z = np.cos(t)

fig = plt.figure()
ax = plt.axes(xlim=(0, twopi*nper), ylim=(-1, 1))

# Non-animated.
ax.plot(t, z, 'k')
ax.set_xlabel(r'$\omega t$')

# Animated.
line, = ax.plot([], [], 'k:o', alpha=0.5)
text = ax.text(1, 1.05, '')
line.set_data([], [])

def init():
return line, text

def animate(i):
# Create a sampling (ts and zs) with decreasing frequency.
ts, zs = [], []
di = twopi / (nper*n) * nper * i
ts = di * k
zs = np.cos(ts)
line.set_data(ts, zs)
if di / np.pi< 1:
text.set_text(r'$\omega$ $\Delta t / \pi$ = %1.3f : no aliasing' % (di/np.pi))
else:
text.set_text(r'$\omega$ $\Delta t / \pi$ = %1.3f : ALIASING' % (di/np.pi))
return line, text

return animation.FuncAnimation(fig, animate, init_func=init,
frames=n, interval=100)

aliasanim(n=300, nper=2)

Out[2]:

Once Loop Reflect