Python - Hands On Topics

Installing Python Modules

Follow along with this static notebook to learn how to install non-standard python packages. Remember for for Lonestar 5, you need to do the following:

module load python
pip --trusted-host pypi.python.org install --user moduleyouneed

Interacting with your Data

Python has several libraries for building Graphical User Interfaces, including Tkinter and wxPython. Jupyter also has ways to interact with the data, by using Widgets. Working with the inflammation dataset, can you update the script to include a scroll bar that allows you to select which datasets are graphed?

Solution

First make sure you are in the directory with the data.

%matplotlib inline

import numpy
import glob
import matplotlib.pyplot
from ipywidgets import interact

# make sure you are already in the directory with the data
filenames = sorted(glob.glob('inflammation*.csv'))

# create a new list to hold all the data
all_data=[]

# for each data file, append the contents to the data list
for f in filenames:
    print(f)
    all_data.append(numpy.loadtxt(fname=f, delimiter=','))

# use widgets to select the array
@interact(x=(0,11,1))
def plot_mean_max_min(x=0):
    fig = matplotlib.pyplot.figure(figsize=(10.0, 3.0))

    axes1 = fig.add_subplot(1, 3, 1)
    axes2 = fig.add_subplot(1, 3, 2)
    axes3 = fig.add_subplot(1, 3, 3)

    axes1.set_ylabel('average')
    axes1.plot(numpy.mean(all_data[x], axis=0))

    axes2.set_ylabel('max')
    axes2.plot(numpy.max(all_data[x], axis=0))

    axes3.set_ylabel('min')
    axes3.plot(numpy.min(all_data[x], axis=0))

    fig.tight_layout()
    matplotlib.pyplot.show()

Pandas - The Python Data Analysis Library

The Pandas library is an excellent option for working with tabular datasets. If you plan to process and visualize data of this type using Python, it is a good idea to learn more about using Pandas. Data Carpentry has an excellent tutorial on analysing tabular data, using an ecology dataset as the example.

Faster!

Try making multi_process.py run even faster. Use any technique, but the same 400,000,000 multiplications must be calculated.

Biopython

Explore the Biopython Tutorial and Cookbook. Many functions have been created to assist with common life science computing tasks. See anything that could help you?