coursera-ml005-unittests
Formal unit tests for Coursera Machine Learning ml-005 instructed by Andrew Ng
To run the tests, follow these steps:
-
clone the git repository to your machine. Here I'll use 'f:/ml005ut':
git clone git@github.com:cod3monk3y/coursera-ml005-unittests.git f:/ml005ut
-
launch Octave and add the directory for the current assignment to your load path. For example, ex2 (Logistic Regression) would use:
>> addpath('f:/ml005ut/ex2/')
-
run the tests!
>> test test_ex2
PASSED 6 out of 6 tests
Conventions and Guidelines
- vectors are represented as column vectors unless they are required to be row vectors by the submit script.
- all unit tests that I submit here are passing, and the code I tested against these tests has been accepted by the submit script.
- names of unit tests will be obfuscated/redacted to hide the nature of the test, so that they don't provide overly detailed hints towards implementation. See discussion on Honor Code here: https://class.coursera.org/ml-005/forum/thread?thread_id=1294
Adding your own tests
Feel free to fork this project and send me pull requests!
To add your own tests, create a new function like test_foo.m
and add a line to test_ex2.m
:
%!test test_foo()
Please do not commit any nonworking unit tests!!!
Please obfuscate or redact your test names. See guidelines above re: Honor Code.
Thanks, and I hope this helps everyone out!