Talk:Jaro similarity: Difference between revisions
no edit summary
m (Markjreed moved page Talk:Jaro distance to Talk:Jaro similarity: Described task calculates the similarity (1=identical) rather than the distance (0=identical)) |
Not a robot (talk | contribs) No edit summary |
||
| (One intermediate revision by one other user not shown) | |||
Line 14:
This page seems to be talking about the Jaro ''similarity'' (0=completely distinct, 1=identical), rather than the Jaro ''distance'' which is the other way around (0=identical, 1=maximally far apart). Propose that the task be renamed and the description reworded as such. --[[User:Markjreed|Markjreed]] ([[User talk:Markjreed|talk]]) 17:36, 30 May 2021 (UTC)
== Wrong last digit ==
Some Outputs show<br>
0.9444444444444445<br>
0.7666666666666666<br>
instead of<br>
0.9444444444444444<br>
0.7666666666666667<br>
Why is tat so?? --Walter Pachl 20:13, 20 December 2021 (UTC)
: I'd wager it's just floating point inaccuracy. Not all languages handle them the exact same way. Operation order also matters: floating point addition is not commutative. Given that, slight variations are probably to be expected. --[[User:Not a robot|Not a robot]] ([[User talk:Not a robot|talk]]) 23:51, 20 December 2021 (UTC)
| |||