OpenEuphoria

Re: data analysis

• Posted by Graeme <graemeburke at CROSSWINDS.NET> Jan 17, 2001
• 646 views

--=====================_979651966==_

At 08:44 AM 17/01/01 +1100, you wrote:
>At 02:14 PM 16/01/01 -0800, you wrote:
>>Kat wrote:
>>
>>> Graeme, since you did that amazing
>>> pattern finding code, i was wondering
>>> what your thoughts would be on this problem...
>>
>>I'm obviously not Graeme, but I couldn't resist. This routine will compare
>>two strings, returning the differences between them. For example:
>>
>
>>    return result
>>
>>end function
>>
>
>Bah! Got me. mine's still about half a hour away. :)
>(Just woke up), I have a feeling it might be a fair
>bit quicker than that one, though.
>
>Graeme.





OK so here it is:

I've tested David's and my own solution with
the following test data:

{"this","that"},
{"actually","actaully"},
{"Pittsborough","Pittsburg"},
{"Pittsberg","Pittsburg"},
{"shello","hellos"},
{"fhfellos","shello"}

results as follows

( p=pass , f=fail )

                DC      GB

test1           p       p
test2           p       p
test3           p       p
test4           f       p
test5           p       f
test6           f       f

Depending on how harsh you're feeling, test3
could also be considered a failure for both.
('r' not matched)

Benchmarking indicates that my routine
is about 4.5 x faster than David's (on a P3)

( hopefully this will prompt David to write a
routine using match() that is faster and better
than both of the above )


routines & test prog attached

Graeme















--=====================_979651966==_
 x-mac-type="705A4950"; x-mac-creator="705A4950"