A Python based battery of cognitive psychology tasks.
I needed to quickly put together a battery for some of my experiments. Currently includes a few tasks that I use for my own projects, but the battery was designed in a modular way so it’s easy to extend it with other tasks in the future.
It allows you to choose which tasks you want to administer, and in which order. Currently it saves results to an Excel file in multiple sheets, but I’m considering moving this over to .csv in the future.
Check the Github repo readme for more details on usage.
Python package containing tools to calculate gait parameters from accelerometer signals. Includes modules for signal processing, peak detection, and gait parameter estimation from time-series data. Additionally this package can be used to estimate physical activity counts from sensor data.
This was originally created as a library of tools for my smartphone-based projects as I needed a reusable set of tools for extracting gait parameters from acceleration signals.
Python/Theano implementations of various machine learning algorithms.
This code base was created for a course on computational modelling and contains Theano implementations of different machine learning algorithms (e.g. artificial/convolutional neural networks, logistic regression).
The goal was to document my learning process, but this repo also serves as an implementation reference for training various models on a GPU.
Languages: Java (Android)
For a number of my research projects I needed to record an individuals vertical acceleration while they were walking. To do that, I developed an Android application to that records sensor data to a SQLite database. On GitHub I currently have the one-phone variant of that application.
So you have 500 students in one of your classes and they just had a midterm exam filled with multiple choice questions. How to quickly grade their MC exams? Use this. It takes a scantron .txt file as input, calculates each student’s total and average scores, calculates the classes average and provides some other descriptive statistics (SD, range etc.), and outputs the data to a couple of files. I have also added an option to drop certain questions from the calculations.
All you need to do is edit the code by changing 1) the locations of the various files, and 2) change the existing answer list with the correct answers for your test. Scantron exams are messy so you may have to double check the scantron output (.txt) files before running this (e.g. check for students incorrectly filling out their name, not filling in a bubble enough etc.).
This was written specifically for scantron machines that output txt files in a specific format. The code can easily be adjusted to read whatever format your scantron file outputs.