Add the one-parameter Lindley distribution #1678
Conversation
ararslan
added
the
feature: new distribution
Codecov ReportBase: 85.71% // Head: 85.40% // Decreases project coverage by
Additional details and impacted files@@ Coverage Diff @@
## master #1678 +/- ##
==========================================
- Coverage 85.71% 85.40% -0.31%
==========================================
Files 130 131 +1
Lines 8230 8286 +56
==========================================
+ Hits 7054 7077 +23
- Misses 1176 1209 +33
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
ararslan
force-pushed
the
aa/lindley
branch
2 times, most recently
from
6ede24e to
10af11d
Compare
| return Lindley{typeof(θ)}(θ) | ||
| end | ||
|
|
||
| Lindley(θ::Integer; check_args::Bool=true) = Lindley(float(θ); check_args=check_args) |
There was a problem hiding this comment.
Is this method needed? For most/all functions integers should work equally well, I assume?
| Lindley(θ::Integer; check_args::Bool=true) = Lindley(float(θ); check_args=check_args) |
There was a problem hiding this comment.
Originally I didn't have this but one of the automated tests checks that integer input gives you a float parameter.
[ci skip] Co-authored-by: David Widmann <devmotion@users.noreply.github.com>
This reverts commit f45e622.
|
I really appreciate the thorough review, @devmotion! |
|
(Note that one of the commits here includes |
|
Thanks again! |
This requires a new dependency on LambertW.jl, which is actively maintained and has no dependencies of its own. I'm also currently usingevalpoly, which was introduced in Julia 1.4, but this package currently requires at least 1.3. IMO we should raise the minimum supported version to 1.6 anyway; I've done so in a separate commit.