This project focuses to improve yt's test framework. At present, yt's Python code coverage is only 25% (unit and answer testing) and the test runtime is approximately 45 minutes. The aim of this project is to increase code coverage and reduce test runtime.

First, I propose the use of Coveralls, which is a tool to monitor the code coverage and is free for open source repositories. This would not only help in analyzing the key areas that need immediate attention for coverage but will also help in maintaining higher code coverage for future developments.

yt's test suite could be divided into three areas, namely, Python unit tests, Cython test cases and answer testing. I will enhance the yt test suite by writing test cases for the flows that are not being tested currently. Runtime of tests could be improved by optimizing (or reducing) answer testing and image comparisons tests for visualization and volume rendering modules. This project also focuses on streamlining test cases for different geometries and data styles to improve the runtime of tests. Furthermore, the runtime of test suites varies on Linux and OSX, thereby giving us a scope of improvement.

Organization

Student

Abhishek Singh

Mentors

  • Kacper Kowalik
  • Nathan Goldbaum
  • Colin Marc
close

2018