in reply to Re^4: A better implementation of LCSS? (Memoize)
in thread A better implementation of LCSS?
Memoize() is only of benefit if you call the memoised function multiple times with the same arguments -- ie. when it can return the previously returned value instead of recalculating it.
Repeating the same test for a benchmark of a lcss() function makes for a totally artificial test.
Of course, you could also apply the memoisation to the XS version and you'd restore the differential.
|
|---|