So what your saying is that my estimate of the performance of your Trie implementation is wildly optimistic.
Therefore the number of comparisons/second you achieved is less than the 21,250,000 figure I used as a divider in the rest of my calculations.
Ergo: Your proposal would take even longer to run.
In reply to Re^5: Fuzzy Searching: Optimizing Algorithm Selection
by BrowserUk
in thread Fuzzy Searching: Optimizing Algorithm Selection
by Itatsumaki
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |