Target Markdown
In worcs: Workflow for Open Reproducible Code in Science

knitr::opts_chunk$set(collapse = TRUE, comment = "#>")
library("worcs")
# We recommend that you prepare your raw data for analysis in 'prepare_data.R',
# and end that file with either open_data(yourdata), or closed_data(yourdata).

{boilerplate}

Target Markdown

Target Markdown is a powerful R Markdown interface for reproducible analysis pipelines, and the chapter at https://books.ropensci.org/targets/markdown.html walks through it in detail. This R Markdown report the example from the chapter. Try it out in both interactive and non-interactive modes, either by running the code chunks in different ways or setting the tar_interactive chunk option.

Packages

The example requires targets version 0.5.0.9000 or above.

Setup

Near the top of the document, you may wish to remove the _targets_r directory previously written by non-interactive runs of the report. Otherwise, your pipeline may contain superfluous targets.

library(targets)
tar_unscript()

Globals

We first define some global options/functions common to all targets.

```{targets example-globals, tar_globals = TRUE} options(tidyverse.quiet = TRUE) tar_option_set(packages = c("worcs"))

# Targets

Our first target borrows the `airquality` dataset built into base R.

```{targets raw-data}
tar_target(raw_data, airquality)

Alternatively, you can use a worcs dataset, which you have saved in the prepare_data.R script by running running the following code, with either closed_data() or open_data():

open_data(airquality)

You may then load it like so:

```{targets worcs-data} tar_target(worcs_data, load_data(to_envir = FALSE)$airquality)

Our next targets preprocess the data, make a histogram, and fit a model.

```{targets downstream-targets}
list(
  tar_target(data, worcs_data[complete.cases(worcs_data), ]),
  tar_target(dens_data, density(data[["Ozone"]]))
)

Set the tar_simple chunk option to TRUE to define a single target with the command in the code chunk. The chunk below only contains lm(Ozone ~ Wind + Temp, data) in the source, but because tar_simple is TRUE, it is shorthand for tar_target(name = fit, command = lm(Ozone ~ Wind + Temp, data)). All other arguments to tar_target() are set to their default values (configurable with tar_option_set()).

```{targets fit, tar_simple = TRUE} lm(Ozone ~ Wind + Temp, data)

# Pipeline

If you ran all the `{targets}` chunks in non-interactive mode, then your R scripts are set up to run the pipeline.

```r
tar_make()

Output

You can retrieve results from the _targets/ data store using tar_read() or tar_load().

tar_read(fit)

plot(tar_read(dens_data))

The targets dependency graph helps your readers understand the steps of your pipeline at a high level.

tar_visnetwork()

At this point, you can go back and run {targets} chunks in interactive mode without interfering with the code or data of the non-interactive pipeline.