flextable R package
Documentation: https://ardata-fr.github.io/flextable-book/
The flextable package provides a framework for easily create tables for
reporting and publications. Tables can be easily formatted with a set of
verbs such as bold()
, color()
, they can receive a header of more
than one line, cells can be merged or contain an image. The package make
it possible to build any table for publication from a `data.frame’.
An API is available to let R users create tables for reporting and
control their formatting properties and their layout. A flextable
object is a data.frame representation, it can be manipulated with
functions that give control over:
- header, body and footer content
- text, paragraphs, cells and border formatting of any element
- displayed values
Installation
install.packages("flextable")
You can get the development version from GitHub:
devtools::install_github("davidgohel/flextable")
Resources
Documentation
The help pages are in a bookdown located at: https://ardata-fr.github.io/flextable-book/
Manuals are available at: https://davidgohel.github.io/flextable/reference/index.html
Getting help
This project is developed and maintained on my own time. If you have
questions about how to use the package, visit Stackoverflow and use tags
flextable
and r
Stackoverflow
link! I
usually read them and answer when possible.
Contributing to the package
Code of Conduct
Anyone getting involved in this package agrees to our Code of Conduct.
Bug reports
When you file a bug report, please spend some time making it easy for me to follow and reproduce. The more time you spend on making the bug report coherent, the more time I can dedicate to investigate the bug as opposed to the bug report.
Contributing to the package development
A great way to start is to contribute an example or improve the documentation.
If you want to submit a Pull Request to integrate functions of yours, please provide:
- the new function(s) with code and roxygen tags (with examples)
- a new section in the appropriate vignette that describes how to use the new function
- corresponding tests in directory
tests/testthat
.
By using rhub (run rhub::check_for_cran()
), you will see if everything
is ok. When submitted, the PR will be evaluated automatically on travis
and appveyor and you will be able to see if something broke.