Open
Conversation
This distribution is tricky to implement as there is no closed form for most functions. PDF is computed based on `wrightbessel`. CDF and quantile are computed numerically. Performance is similar to reference R package tweedie. When `1 < p < 2` it is continuous but with a point mass at zero so it is not completely correct to consider it a `ContinuousUnivariateDistribution`. When `p = 1` it is even discrete.
Member
|
Tests fail currently. |
devmotion
reviewed
Mar 23, 2026
Comment on lines
+54
to
+56
| @check_args Tweedie (μ, μ >= zero(μ)) | ||
| @check_args Tweedie (σ, σ >= zero(σ)) | ||
| @check_args Tweedie (p, 1 <= p <= 2) |
Member
There was a problem hiding this comment.
Not 100% sure but IIRC you can use a single check_args with a Tuple of checks.
Comment on lines
+63
to
+64
| Tweedie(μ::Integer, σ::Integer, p::Integer; check_args::Bool=true) = | ||
| Tweedie(float(μ), float(σ), float(p); check_args=check_args) |
Member
There was a problem hiding this comment.
Not needed?
Suggested change
| Tweedie(μ::Integer, σ::Integer, p::Integer; check_args::Bool=true) = | |
| Tweedie(float(μ), float(σ), float(p); check_args=check_args) |
| Tweedie(float(μ), float(σ), float(p); check_args=check_args) | ||
|
|
||
| #### Conversions | ||
| convert(::Type{Tweedie{T}}, μ::S, σ::S, p::S) where {T <: Real, S <: Real} = Tweedie(T(μ), T(σ), T(p)) |
Member
There was a problem hiding this comment.
IMO no such method should be defined - users should use the constructor:
Suggested change
| convert(::Type{Tweedie{T}}, μ::S, σ::S, p::S) where {T <: Real, S <: Real} = Tweedie(T(μ), T(σ), T(p)) |
| #### Parameters | ||
|
|
||
| params(d::Tweedie) = (d.μ, d.σ, d.p) | ||
| partype(d::Tweedie{T}) where {T} = T |
Member
There was a problem hiding this comment.
Suggested change
| partype(d::Tweedie{T}) where {T} = T | |
| partype(::Tweedie{T}) where {T} = T |
| #### Statistics | ||
|
|
||
| mean(d::Tweedie) = d.μ | ||
| var(d::Tweedie) = d.σ^2 * d.μ^d.p |
Member
There was a problem hiding this comment.
I'd suggest defining std as well to avoid unnecessary squaring and subsequent square root calculations:
Suggested change
| var(d::Tweedie) = d.σ^2 * d.μ^d.p | |
| var(d::Tweedie) = d.σ^2 * d.μ^d.p | |
| std(d::Tweedle) = d.σ * d.μ^(d.p/2) |
Comment on lines
+5498
to
+5506
| "properties": { | ||
| "location": 2, | ||
| "scale": 1.5, | ||
| "shape": 1.01, | ||
| "mean": 2, | ||
| "var": 4.53129997525524, | ||
| "skewness": 1.07498594232179, | ||
| "kurtosis": 1.16703630862699 | ||
| }, |
Member
There was a problem hiding this comment.
Suggested change
| "properties": { | |
| "location": 2, | |
| "scale": 1.5, | |
| "shape": 1.01, | |
| "mean": 2, | |
| "var": 4.53129997525524, | |
| "skewness": 1.07498594232179, | |
| "kurtosis": 1.16703630862699 | |
| }, |
Comment on lines
+5531
to
+5539
| "properties": { | ||
| "location": 2, | ||
| "scale": 1.5, | ||
| "shape": 1.5, | ||
| "mean": 2, | ||
| "var": 6.36396103067893, | ||
| "skewness": 1.89201693432086, | ||
| "kurtosis": 4.7729707730092 | ||
| }, |
Member
There was a problem hiding this comment.
Suggested change
| "properties": { | |
| "location": 2, | |
| "scale": 1.5, | |
| "shape": 1.5, | |
| "mean": 2, | |
| "var": 6.36396103067893, | |
| "skewness": 1.89201693432086, | |
| "kurtosis": 4.7729707730092 | |
| }, |
Comment on lines
+5564
to
+5572
| "properties": { | ||
| "location": 2, | ||
| "scale": 1.5, | ||
| "shape": 1.99, | ||
| "mean": 2, | ||
| "var": 8.93783245893332, | ||
| "skewness": 2.97467268454119, | ||
| "kurtosis": 13.2507835119916 | ||
| }, |
Member
There was a problem hiding this comment.
Suggested change
| "properties": { | |
| "location": 2, | |
| "scale": 1.5, | |
| "shape": 1.99, | |
| "mean": 2, | |
| "var": 8.93783245893332, | |
| "skewness": 2.97467268454119, | |
| "kurtosis": 13.2507835119916 | |
| }, |
Comment on lines
+5597
to
+5605
| "properties": { | ||
| "location": 2, | ||
| "scale": 1.5, | ||
| "shape": 2, | ||
| "mean": 2, | ||
| "var": 9, | ||
| "skewness": 3, | ||
| "kurtosis": 13.5 | ||
| }, |
Member
There was a problem hiding this comment.
Suggested change
| "properties": { | |
| "location": 2, | |
| "scale": 1.5, | |
| "shape": 2, | |
| "mean": 2, | |
| "var": 9, | |
| "skewness": 3, | |
| "kurtosis": 13.5 | |
| }, |
Member
There was a problem hiding this comment.
We should also test Float32 inputs, integer inputs and mixes of Float32, Float64, and integers, and add @inferred around the function calls. This would have caught the type stability issues.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This distribution is tricky to implement as there is no closed form for most functions. PDF is computed based on
wrightbessel(JuliaMath/SpecialFunctions.jl#512). CDF and quantile are computed numerically from PDF. Performance is similar to reference R package tweedie.When
1 < p < 2it is continuous but with a point mass at zero so it is not completely correct to consider it aContinuousUnivariateDistribution. Whenp = 1it is even discrete. However I wasn't sure that justifies adding a new type likeSemiContinuousUnivariateDistribution.This distribution is useful to allow estimating Tweedie models in GLM.