FeatureTable

Note: I still consider FeatureTable to be in the pre-release stage. The code itself is well tested and I use it in my own research, but the API has not stabalized and is still subject to breaking changes when I cut a new release.

Installing

The simplest was is probably by using the remotes package:

remotes::install_github(
  repo = "mooreryan/featuretable",
  build_vignettes = TRUE, build = TRUE, force = TRUE, upgrade = "never",
  # This will ensure that all suggested dependencies are installed as well.
  dependencies = TRUE
)

Note: you can install a specific tag or relase by replacing this "mooreryan/featuretable" with something like this "mooreryan/featuretable@0.0.11". (That would install tag 0.0.11.)

Alternatively, you could procure the code, and install with devtools. To obtain the code, either:

• Go to the Releases page and click on the latest release,
• Or, if you want to follow the main branch, clone this repository.

Then install with devtools. (Ensure that you are running this from the root of the project's source directory.)

devtools::install(
  build_vignettes = TRUE, build = TRUE, force = TRUE, upgrade = "never",
  # This will ensure that all suggested dependencies are installed as well.
  dependencies = TRUE
)

Dependencies

Here is some info about the dependencies. Note that if you used one of the code snippets above, all the optional dependencies will be installed as well.

Required

These packages must be installed to use featuretable.

• R6
• rlang

Suggested

• If you want nice plots, I suggest you also install ggplot2, ggrepel, and biplotr.
• If you want fancy zero replacement, you should install zCompositions.
• If you need to import or export phyloseq objects, then you need to install phyloseq.
• If you want to build the vignettes yourself, you will need knitr and rmarkdown. Also, some of the vignettes use magrittr to keep things clean.

Test suite dependencies

• If you want to run the test suite locally, then you need to install testthat.
• Some of the tests require additional packages. If you don't have them, those tests should be skipped.
  • Testing plotting functions requires vdiffr.
  • If you want to run the coverage report for yourself, then you will need to install covr.
  • Some of the tests compare bespoke implementations of certain functions to their counterparts in dplyr, tibble, and tidyr.
  • Property tests require the hedgehog package.

Usage

Vignettes

FeatureTable comes with some detailed vignettes describing its usage. After installing, you can view them with the following command:

browseVignettes("featuretable")

Function docs

Additionally, each function has a lot of info in the help section. You can run ?featuretable to get an overview of everything. Click on the links in there to get to individual function docs. You can also access function docs with the ?, e.g., ?keep_samples.

R v3 and v4 compatibility

R CMD CHECK, which includes unit tests, are run via GitHub CI for the following versions:

• MacOS, Windows, & Ubuntu
  • Latest release
• Ubuntu
  • 4.1.3
  • 4.0.5
  • 3.6.3
  • 3.5.3

So as long as that is passing (see the badge at the top of the README), then you can be (pretty) sure the package works with both R v3 and v4.

Notes:

• Each time data.frame or as.data.frame is called in the FeatureTable code and tests , stringsAsFactors = TRUE is passed in as an argument.
  • So, if you need a specific stable sorting of your string-like data, you need to set the factor levels yourself.
  • In theory, you should be fine passing in data.frames that don't have their strings as factors, since FeatureTable will convert all strings to factors, but I should probably add tests for this.
  • This behavior may change in the future though!
• One thing to note is that as.data.frame.FeatureTable does not have a stringsAsFactors parameter. This shouldn't be a problem as string-esque data is not allowed in the data field anyway.

Hacking

Notes for working on the code.

• If you make an update to the core R6 class, make sure you regenerate the .rda files in data, using the scripts in data-raw.
  • Note that generating the data files requires the DivNet R package.
  • You can use the recipe in the justfile for this.

Just

Just go and install just. Then use the recipes in the justfile.

Default branch is now main

The default branch is now main, and the master branch no longer exists.

If you have a local clone using master as the default branch, you can update it by running the following commands.

git branch -m master main
git fetch origin
git branch -u origin/main main
git remote set-head origin -a

Optionally, run the following command to remove tracking references to the old branch name.

git remote prune origin