A String Search algorithm I wrote is performing incredibly well in my own test, but I'd like to test it more rigorously to know how well it objectively performs. Anyone know of a resource I can use to run such tests?
The test I created puts together 5000 random words, then randomly picks one of those words as the 'pattern'. When I run my test 1000 times (each time with a new set of 5000 random words and a new 'pattern'), my algorithm beats (in speed) Boyer-Moore roughly 1/3 of the times, and beats/equals Boyer-Moore roughly 3/5 of the times.
When I take the average time of all 1000 tests, sometimes my algorithm has a better average, but more often Boyer-Moore has a better average, though usually not by much (<0.0005ms).
Here's how the comparison looks as a scatter plot for n=1000.The black lines compare the two algorithms for any specific set (blue marks are my algorithm times, orange are for Boyer-Moore). The red line compares the two averages over these n tests. Averages are written in the title.
So if you know of any resource that can help me test this, please send my way.
Thanks.