[
https://issues.apache.org/jira/browse/NUMBERS-77?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17329363#comment-17329363
]
Gilles Sadowski commented on NUMBERS-77:
----------------------------------------
bq. In your diff [...]
Let's refer to what's in "master" (there were fixes).
bq. comparisons break down when NaN is input. Is this the behaviour we want?
Probably not if it can be helped. ;-)
bq. I like having the term "precision" in the name [...]
Then why
bq. prefer if the interface were top-level
since the nesting offers both?
bq. I'm going to need to update about a bajillion variable names [...]
Why?
It's perfectly fine that
{code}
final DoublePrecisionContext precision = new
EpsilonDoublePrecisionContext(1e-4);
{code}
becomes
{code}
final Precision.DoubleEquivalence precision =
Precision.doubleEquivalenceOfEpsilon(1e-4);
{code}
bq. [...] values [0.01, 0.02, 1, 2] are "correctly" sorted when using an
epsilon value of 0.1 but so are [0.02, 0.01, 1, 2].
Not a problem since indeed 0.01 and 0.02 are _equivalent_. Hence the name. ;-)
> Move utilities from "Commons Geometry"
> --------------------------------------
>
> Key: NUMBERS-77
> URL: https://issues.apache.org/jira/browse/NUMBERS-77
> Project: Commons Numbers
> Issue Type: Task
> Reporter: Gilles Sadowski
> Priority: Major
> Fix For: 1.1
>
> Attachments: NUMBERS-77.diff
>
> Time Spent: 1h 10m
> Remaining Estimate: 0h
>
> "Commons Geometry" defines utilities that would be a better fit in this
> component.
> Duplication of general-purpose codes should be avoided, in order to benefit
> from consolidated usage (bug reporting, performance enhancements, ...).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
