Data analysis and experiment control with Python

2020 Winter

Date: Thursday, 11:00

First class on January 16

We'll find an extra class for the cancelled class on the week of the BCCCD.

Location: Budapest, Október 6 street 7, room 103

Course description

The aim of the course is to learn basic experiment control, data analysis and simulation in Python language. The sessions include both lectures and hands-on class works. The course supposes that you already know the basics of Python language.

Learning outcomes

By the end of this course, students will be able to:

• • Create experiment control Python script for most of the experiment designs

  • Perform classic behavioral data analysis

  • Run some of the advanced data analysis methods

  • Start to learn and use data analyses specific to some non-behavioral methods (e.g., EEG, fMRI)

  • Run simple simulations

Course requirements

For both requirements any topic might be chosen that is appropriate according to the descriptions below. Specific topics should be approved by the lecturer.

Experiment control script (40% of the final grade; deadline for topic approval: Feb 13, deadline for the completed script: Feb 27, deadline for code comments: Mar 19)

• • Write your own experiment, preferably in PsychoPy (the script should not generated by the Builder view). (Let me know if you'd use another system.)

  • Comment the code written by others. Go to https://gitlab.com/krajcsi/python-class-ceu-2020-winter/ > Repository > Commits > Choose the code maker seen in the commit message > then read and comment the code (click on the callout icon at the begenning of a row). Note that some files should be separately opened/displayed in the commit view. If you don't have anything to add for any of the files in a commit, then add a comment at the bottom of the page to let us know that you don't have any comments.

Data analysis and/or simulation (60% of the final grade; deadline for topic approval: Mar 12, deadline for the completed script: Apr 2)

• Choose one of the following options:

  • • Analyze your own research data with Python.

    • Run your own simulation in Python.

    • Code some new function for CogStat.

• • Comment the code written by others at https://gitlab.com/krajcsi/python-class-ceu-2020-winter/ .

Course schedule

(Some of the links lead to Hungarian pages or slides at the moment. Expect some updates later.)

(Based on the interest of the group and based on some feedback expect some changes in the course schedule.)

Main topics:

• • Catching up on Python

  • Why to use Python?

    • Review of basic elements in Python (slides and slides)

  • Experiment control with computer

    • General considerations - constrains and problems to solve

    • OpenSesame - see the structure of the elements without the code

  • PsychoPy

• Data analysis in Python (slides)

  • • Jupyter Notebook and Jupyter Lab (slides)

    • Some of the relevant standard modules

      • math, random, string, time, timeit

• • Data handling and base functions

    • numpy, scipy

  • Plotting the data

    • matplotlib

  • Statistics

    • pandas, statistics, scipy.stats, statsmodels, CogStat (slides), rpy2

    • Specific statistical methods: Monte Carlo, bootstrapping (slides)

    • Python tools for specific cognitive methods: EEG, fMRI

• Any other wishes?

Week by week schedule:

Recommended reading

Documentations and tutorials of the used language and software.