exercises.md

Exercises for Day 4

Testing and documentation of code, plotting with matplotlib

1. Testing code with py.test

a. Install pytest using e.g.

pip install pytest --user

b. Write a module test_simple_math.py which tests the correctness of the math operations inside simple_math.py.

You can run the tests using the py.test command. Note that the name of your test functions should start with test_.

2. Documenting code

a. Write some meaningful docstrings for the simple_math.py module, e.g. following the numpy_doc standard.

b. Create a documenation html page for the simple_math.py module using Sphinx.

Follow the instructions from the lecture notes.

3. Plotting with matplotlib

a. Download the jupyter notebook customized_plotting.ipynb and try to understand the individual steps in creating a plot and customizing it

b. Produce your own data (curves, images, point clouds, etc...) and create a nice plot. Get inspired by the matplotlib gallery and use the extensive documentation of the pyplot API

4. Optimization with SciPy

Fitting Datasets

In the repository there's a file, I_q_IPA_exp.npy, with an experimental set of points giving the scattering strength of a sample as a function of the scattering vector. You also have, I_q_IPA_model.npy, a set of points produced by a theoretical model that also give the scattering strength as a function of the scattering vector. You would like to scale the points from the theoretical model to give the best possible fit to the experimental dataset. Use numpy.load() to load the arrays. For both arrays [:,0] correspond to the scattering vector and [:,1] corresponds to the scattering strength. A few useful SciPy functions might be scipy.optimize.minimize_scalar() and scipy.interpolate.interp1d(). Matplotlib might be helpful to visualize the result, e.g. with plot() or scatter().