epitargets

epitargets extends the {targets} pipeline framework with automatic date-tracking and age-based cues, and other convenience functions.

Installation

install.packages(
   "epitargets",
   repos = c("https://economic.r-universe.dev", getOption("repos"))
)

Usage

library(epitargets)

Track when targets were last built

tar_target_date() is a drop-in replacement for targets::tar_target(). It creates your target plus a companion _date target that records when it ran:

targets::tar_dir({
  targets::tar_script({
    library(epitargets)
    list(
      tar_target_date(wages, data.frame(year = 2020:2024, wage = c(15, 15, 16, 17, 18)))
    )
  })
  targets::tar_make()
  print(targets::tar_read(wages))
  print(targets::tar_read(wages_date))
})
#> + wages dispatched
#> ✔ wages completed [0ms, 191 B]
#> + wages_date dispatched
#> ✔ wages_date completed [0ms, 81 B]
#> ✔ ended pipeline [124ms, 2 completed, 0 skipped]
#>   year wage
#> 1 2020   15
#> 2 2021   15
#> 3 2022   16
#> 4 2023   17
#> 5 2024   18
#> [1] "2026-07-04"

Auto-refresh targets on a schedule

tar_age_date() adds an age-based cue so the target automatically re-runs after a time period. Useful for API calls or any data that goes stale:

targets::tar_dir({
  targets::tar_script({
    library(epitargets)
    list(
      tar_age_date(daily_data, Sys.time(), age = as.difftime(1, units = "days"))
    )
  })
  targets::tar_make()
  print(targets::tar_read(daily_data))
  print(targets::tar_read(daily_data_date))
})
#> + daily_data dispatched
#> ✔ daily_data completed [0ms, 96 B]
#> + daily_data_date dispatched
#> ✔ daily_data_date completed [0ms, 81 B]
#> ✔ ended pipeline [127ms, 2 completed, 0 skipped]
#> [1] "2026-07-04 15:41:27 EDT"
#> [1] "2026-07-04"

Summarize freshness across targets

collect_target_date() gathers _date targets into a single tibble so you can see at a glance when each target last ran:

targets::tar_dir({
  targets::tar_script({
    library(epitargets)
    list(
      tar_target_date(wages, data.frame(wage = 18)),
      tar_age_date(prices, data.frame(price = 3.50)),
      targets::tar_target(freshness, collect_target_date(wages_date, prices_date))
    )
  })
  targets::tar_make()
  targets::tar_read(freshness)
})
#> + prices dispatched
#> ✔ prices completed [1ms, 118 B]
#> + wages dispatched
#> ✔ wages completed [1ms, 118 B]
#> + prices_date dispatched
#> ✔ prices_date completed [0ms, 81 B]
#> + wages_date dispatched
#> ✔ wages_date completed [0ms, 81 B]
#> + freshness dispatched
#> ✔ freshness completed [2ms, 179 B]
#> ✔ ended pipeline [148ms, 5 completed, 0 skipped]
#> # A tibble: 2 × 2
#>   name   time      
#>   <chr>  <date>    
#> 1 wages  2026-07-04
#> 2 prices 2026-07-04

Read a target and keep a handle to it

tar_read_stash() reads a target and, besides returning it, leaves a copy in the global environment as .target so you can keep exploring it without re-running the read:

tar_read_stash(wages)
.target |> dplyr::filter(year == 2024)

It really shines as an RStudio/Positron addin. Put the cursor on a target name in _targets.R or an analysis script and trigger the “Stash target under cursor” addin (bind it to a keyboard shortcut, just like targets::rstudio_addin_tar_read()). The value prints to the console and lands in .target.

The addin extracts the symbol under the cursor with atcursor, an optional dependency. If it isn’t installed the addin will offer to install it for you from its r-universe repository.