Index
InferenceReport.Inference
InferenceReport.PostprocessContext
InferenceReport.ReportOptions
InferenceReport.TargetDescription
InferenceReport.add_markdown
InferenceReport.add_plot
InferenceReport.add_table
InferenceReport.as_doc_page
InferenceReport.default_postprocessors
InferenceReport.gcb_progress
InferenceReport.get_pt
InferenceReport.headless
InferenceReport.lcb
InferenceReport.logz_progress
InferenceReport.moments
InferenceReport.moving_pair_plot
InferenceReport.mpi_standard_out
InferenceReport.pair_plot
InferenceReport.pigeons_inputs
InferenceReport.pigeons_summary
InferenceReport.pigeons_target_description
InferenceReport.pigeons_target_name
InferenceReport.render
InferenceReport.report
InferenceReport.report
InferenceReport.report_to_docs
InferenceReport.reproducibility_info
InferenceReport.round_trip_progress
InferenceReport.swaps_plot
InferenceReport.target_description
InferenceReport.target_name
InferenceReport.trace_plot
InferenceReport.trace_plot_cumulative
InferenceReport.view_webpage
InferenceReport.@reproducible
Types and functions
InferenceReport.Inference
— TypeThe object of a report created by this package:
algorithm
: Algorithm that created the samples (currently assumes it is aPT
struct). Optional, nothing if unspecfied.
chains
: Samples, stored in aChains
struct.
original_dim
: The original dimensionality of chains (the one stored might be truncated via options.max_dim)
InferenceReport.PostprocessContext
— TypeInformation passed to the processors:
inference
: TheInference
on which the report will be based on.
output_directory
: The resolved location of the output report.
generated_markdown
: Vector of Strings used internally to create the markdown page.
generated_dict
: Generated dictionary used internally to create a JSON object for indexing purposes.
options
: TheReportOptions
.
InferenceReport.ReportOptions
— TypeOptions to control report
:
max_moving_plot_iters
: Maximum number of iterations inmoving_pair_plot
. Set to zero to avoid creating such movie.Creating movies are a bit slow, so in interactive mode the default is zero, and 100 in
headless
mode.
target_name
: Name of the target. If nothing, will attempt to auto-detect a name via multiple dispatch of the functionstarget_name
orpigeons_target_name
.
target_description
: Description of the target, specified using aTargetDescription
. Ifnothing
, will attempt to auto-detect it viatarget_description
andpigeons_target_description
.
render
: If true, the report webpage's md files are rendered into html files.
view
: If true, the report webpage will be opened automatically.
postprocessors
: The postprocessors to use. Default isdefault_postprocessors
.
exec_folder
: The directory where the report will be created. A unique folder inresults/all
will be created by default and symlinked inresults/latest
.
reproducibility_command
: A command to instruct users how to reproduce.At the moment implicitly assumes Pigeons will be used, and running these commands should return the Inputs.
writer
: A Documenter.Writer instance used to render the document. Default is an HTML page, but other options including Documenter.LaTeX() are also possible.
show_error_traces
: Useful for debugging, false by default.
max_dim
: Upper bound on the number of dimensions to process. If the number of dimensions exceed this bound, only the prefix of dimensions will be processed.
InferenceReport.TargetDescription
— TypeDescription of a target.
keywords
: Keywords for the target.
bibliography
: Contents of TeXbib
file used to provide bibliographical references in the text, ornothing
if no bibliography should be generated.
text
: Description of the target.
InferenceReport.add_markdown
— Methodadd_markdown(context; title, contents)
InferenceReport.add_plot
— Methodadd_plot(context; file, title, url_help, description, movie)
InferenceReport.add_table
— Methodadd_table(
context;
table,
title,
url_help,
kw_pretty_table_args...
)
InferenceReport.as_doc_page
— Methodas_doc_page(context)
Pass to the pages
argument of makedocs
to properly include the generated page in the Documenter.jl navigation bar.
InferenceReport.default_postprocessors
— Methoddefault_postprocessors()
InferenceReport.gcb_progress
— Methodgcb_progress(context)
Progression of the global communication barrier estimate as a function of the round.
InferenceReport.get_pt
— Methodget_pt(context)
InferenceReport.headless
— Methodheadless(lambda)
Enclose calls to report_to_docs
and report
with different defaults for the view
and render
options, setting them to false by default.
InferenceReport.lcb
— Methodlcb(context)
The local communication barrier estimated in the last round.
InferenceReport.logz_progress
— Methodlogz_progress(context)
Progression of the log-normalization constant estimate as a function of the round.
InferenceReport.moments
— Methodmoments(context)
Table of means and variances.
InferenceReport.moving_pair_plot
— Methodmoving_pair_plot(context)
A PairPlot movie showing the sample in the reference chain moving in the state space.
InferenceReport.mpi_standard_out
— Methodmpi_standard_out(context)
MPI standard output.
InferenceReport.pair_plot
— Methodpair_plot(context)
Create a pair plot using PairPlot.jl.
InferenceReport.pigeons_inputs
— Methodpigeons_inputs(context)
Table of inputs used in Pigeons.
InferenceReport.pigeons_summary
— Methodpigeons_summary(context)
The summary table produced by Pigeons during its execution.
InferenceReport.pigeons_target_description
— Methodpigeons_target_description(target)
By default, dispatch to pigeons_target_description(target, Val(Symbol(pigeons_target_name(target))))
, to handle "vague types" such as TuringLogPotential and StanLogPotential.
InferenceReport.pigeons_target_name
— Methodpigeons_target_name(target)
InferenceReport.render
— Methodrender(context)
InferenceReport.report
— Methodreport(algo_or_chains, options)
InferenceReport.report
— Methodreport(algo_or_chains; args...)
algo_or_chains
can be either a PT
struct (returned by pigeons(..)
) or a Chains
.
args
are used to create a ReportOptions
.
InferenceReport.report_to_docs
— Methodreport_to_docs(algo_or_chains; doc_root, args...)
Similar to report
but with different defaults to facilitate including the result as part of a larger documentation site.
Use in combination with as_doc_page
.
InferenceReport.reproducibility_info
— Methodreproducibility_info(context)
Generated reproducibility script.
InferenceReport.round_trip_progress
— Methodround_trip_progress(context)
Progression of the number of round trips as a function of the round.
InferenceReport.swaps_plot
— Methodswaps_plot(context)
A plot showing progression of mean swap acceptance per chain as a function of round.
InferenceReport.target_description
— Methodtarget_description(context)
InferenceReport.target_name
— Methodtarget_name(context)
InferenceReport.trace_plot
— Methodtrace_plot(context)
Trace plots.
InferenceReport.trace_plot_cumulative
— Methodtrace_plot_cumulative(context)
Cumulative averages as a function of iteration.
InferenceReport.view_webpage
— Methodview_webpage(exec_folder)
InferenceReport.@reproducible
— MacroReturns a pair, where the first argument is the provided expression verbatim, and the second, the expression as a string to be used as the reproducibility_command
argument in report
and report_to_docs
.