Skip to content
This repository has been archived by the owner on Jul 11, 2023. It is now read-only.

PumasAI/PumasPostProcessing.jl

Repository files navigation

PumasPostProcessing

In use to build out Post-Processing functions.

Setup

This package is available through the JuliaPro registry, and can be added in JuliaPro by ]add PumasPostProcessing.

Usage

PumasPostProcessing exposes two APIs - a low level, table generating API and a high-level report generating API.

To use the automated report generation, you can call the exported jmd_report function, which will take in a FittedPumasModel and return a Weave markdown string which can be written to a file.

using Weave
write("report.jmd", PumasPostProcessing.jmd_report(fpm, "My Model Name", "1"))
Weave.weave("report.jmd"; doctype="pandoc2pdf", fig_ext= ".pdf",
pandoc_options = ["-V 'geometry:margin=1in'"], keep_unicode = false)

There is also "dumb", plain-markdown output from to_report_str,as in the following snippet. If you have pandoc installed, there is also a report_to_pdf function, which will automatically create a PDF of the report through LaTeX.

# fpm isa FittedPumasModel
report = to_report_str(fpm)
write("report.md", report)
report_to_pdf("report.pdf", report)

The low-level API consists of functions which take in an FPM, and return tables. These are in the functions param_table, optim_meta_table, and metric_table, which are all exported. They are also documented, and you can view their docstrings in the REPL or JuliaPro documentation browser.

Feature requests and bug reports

Please file any feature requests or bug reports you have as issues on this repo.