[ml] quantizer test harness
Michael C. Toren
mct at toren.net
Mon Aug 31 23:39:13 PDT 2009
On Thu, Aug 27, 2009 at 11:44:30PM -0700, Michael C. Toren wrote:
> Unfortunately, the output is far from identical at the moment :/ I
> created three coordinate test files, wiigee-c/test_data/increment.dat,
> zeros.dat, and repeated.dat. The first two are returning the wrong
> number of elements (8, which happens to be the number of states, rather
> than 20, the number of coordinates in the data file), which hopefully is
> an easy bug to correct. The third file appears to cause the program to
> enter an infinite loop, which is more annoying.
> So, we have some work to do. But, at least we have some metrics to see
> how far along -- and how correct -- we are.
Through judicious use of printf and diff, I found and fixed all the bugs!
There's a new "make test" Makefile target that runs each test_data/*.dat
file through Quantizer.java and quantizer.c and confirms the outputs are
identical. Every test is passing at the moment. :-)
I could use help from others to:
- Confirm the java test suite looks sane and is behaving correctly.
If it isn't, we're confirming our C port against something bogus,
which would utterly and completely suck. Read through the short
wiigee-c/javatest/TestHarness.java, and compare my copies of
Gesture.java and Quantizer.java in wiigee-c/javatest/logic to
those in the upstream wiigee distribution.
- Populate wiigee-c/test_data with more test files. Try to think
of corner cases that may break us.
More information about the ml