diff --git a/dev/api/index.html b/dev/api/index.html index 816b021b..339d84e8 100644 --- a/dev/api/index.html +++ b/dev/api/index.html @@ -7,4 +7,4 @@ x = rand(proposal, 1_000, 100) log_ratios = logpdf.(target, x) .- logpdf.(proposal, x) result = psis(log_ratios) -paretoshapeplot(result)

We can also plot the Pareto shape parameters directly:

paretoshapeplot(result.pareto_shape)

We can also use plot directly:

plot(result.pareto_shape; showlines=true)
source
+paretoshapeplot(result)

We can also plot the Pareto shape parameters directly:

paretoshapeplot(result.pareto_shape)

We can also use plot directly:

plot(result.pareto_shape; showlines=true)
source
diff --git a/dev/index.html b/dev/index.html index 7dd51e36..f626a413 100644 --- a/dev/index.html +++ b/dev/index.html @@ -7,4 +7,4 @@ result = psis(log_ratios)
PSISResult with 1000 draws, 30 chains, and 1 parameters
 Pareto shape (k) diagnostic values:
                    Count       Min. ESS
- (0.5, 0.7]  okay  1 (100.0%)  26838

As indicated by the warnings, this is a poor choice of a proposal distribution, and estimates are unlikely to converge (see PSISResult for an explanation of the shape thresholds).

When running PSIS with many parameters, it is useful to plot the Pareto shape values to diagnose convergence. See Plotting PSIS results for examples.

+ (0.5, 0.7] okay 1 (100.0%) 26838

As indicated by the warnings, this is a poor choice of a proposal distribution, and estimates are unlikely to converge (see PSISResult for an explanation of the shape thresholds).

When running PSIS with many parameters, it is useful to plot the Pareto shape values to diagnose convergence. See Plotting PSIS results for examples.

diff --git a/dev/internal/index.html b/dev/internal/index.html index d7148893..8bcc51eb 100644 --- a/dev/internal/index.html +++ b/dev/internal/index.html @@ -1,2 +1,2 @@ -Internal · PSIS.jl
Edit on GitHub

Internal

PSIS.GeneralizedPareto — Type
GeneralizedPareto{T<:Real}

The generalized Pareto distribution.

This is equivalent to Distributions.GeneralizedPareto and can be converted to one with convert(Distributions.GeneralizedPareto, d).

Constructor

GeneralizedPareto(μ, σ, k)

Construct the generalized Pareto distribution (GPD) with location parameter $μ$, scale parameter $σ$ and shape parameter $k$.

Note

The shape parameter $k$ is equivalent to the commonly used shape parameter $ξ$. This is the same parameterization used by [VehtariSimpson2021] and is related to that used by [ZhangStephens2009] as $k \mapsto -k$.

source
PSIS.fit_gpd — Method
fit_gpd(x; μ=0, kwargs...)

Fit a GeneralizedPareto with location μ to the data x.

The fit is performed using the Empirical Bayes method of [ZhangStephens2009].

Keywords

  • prior_adjusted::Bool=true, If true, a weakly informative Normal prior centered on $\frac{1}{2}$ is used for the shape $k$.
  • sorted::Bool=issorted(x): If true, x is assumed to be sorted. If false, a sorted copy of x is made.
  • min_points::Int=30: The minimum number of quadrature points to use when estimating the posterior mean of $\theta = \frac{\xi}{\sigma}$.
source
+Internal · PSIS.jl
Edit on GitHub

Internal

PSIS.GeneralizedPareto — Type
GeneralizedPareto{T<:Real}

The generalized Pareto distribution.

This is equivalent to Distributions.GeneralizedPareto and can be converted to one with convert(Distributions.GeneralizedPareto, d).

Constructor

GeneralizedPareto(μ, σ, k)

Construct the generalized Pareto distribution (GPD) with location parameter $μ$, scale parameter $σ$ and shape parameter $k$.

Note

The shape parameter $k$ is equivalent to the commonly used shape parameter $ξ$. This is the same parameterization used by [VehtariSimpson2021] and is related to that used by [ZhangStephens2009] as $k \mapsto -k$.

source
PSIS.fit_gpd — Method
fit_gpd(x; μ=0, kwargs...)

Fit a GeneralizedPareto with location μ to the data x.

The fit is performed using the Empirical Bayes method of [ZhangStephens2009].

Keywords

  • prior_adjusted::Bool=true, If true, a weakly informative Normal prior centered on $\frac{1}{2}$ is used for the shape $k$.
  • sorted::Bool=issorted(x): If true, x is assumed to be sorted. If false, a sorted copy of x is made.
  • min_points::Int=30: The minimum number of quadrature points to use when estimating the posterior mean of $\theta = \frac{\xi}{\sigma}$.
source
diff --git a/dev/plotting/index.html b/dev/plotting/index.html index cba6ffe7..525b7b00 100644 --- a/dev/plotting/index.html +++ b/dev/plotting/index.html @@ -11,51 +11,51 @@ plot(result; showlines=true, marker=:+, legend=false, linewidth=2) - + - + - + - + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + -

This is equivalent to calling PSISPlots.paretoshapeplot(result; kwargs...).

+

This is equivalent to calling PSISPlots.paretoshapeplot(result; kwargs...).

diff --git a/dev/search/index.html b/dev/search/index.html index 3a6d4ac7..42650267 100644 --- a/dev/search/index.html +++ b/dev/search/index.html @@ -1,2 +1,2 @@ -Search · PSIS.jl

Loading search...

    +Search · PSIS.jl

    Loading search...