_targets.R

In softloud/pandan: Minimalistic project tracking

library(targets)
# This is an example _targets.R file. Every
# {targets} pipeline needs one.
# Use tar_script() to create _targets.R and tar_edit()
# to open it again for editing.
# Then, run tar_make() to run the pipeline
# and tar_read(summary) to view the results.

# Define custom functions and other global objects.
# This is where you write source(\"R/functions.R\")
# if you keep your functions in external scripts.
summ <- function(dataset) {
  summarize(dataset, mean_x = mean(x))
}

# Set target-specific options such as packages.
tar_option_set(packages =
                 c("tidyverse",
                   "glue",
                   "lubridate"))

# End this file with a list of target objects.
list(
  tar_target(
    raw_obs,
    tibble(component = c("1", "1.1", "1.2", "1.2.1", "1"),) %>%
      mutate(
        project = c(rep("a", length(component) - 1), "b"),

        pred = case_when(
          component == "1.1" & project == "a" ~ "a_1.2.1",
          component == "1" &
            project == "b" ~ "a_1.1"
        ),
        time = sample(1:3, size = length(component), replace = TRUE)
      ) %>%
      select(project, everything())
  ),

  tar_target(raw_proj,
             tibble(
               project = c("a", "b"),
               start_inst = c("2021-05-05", "2021-05-06")
             ) %>%
             mutate(
               start_inst = ymd(start_inst)
             )
             ),

  tar_target(
    obs,
    raw_obs %>%
      mutate(
        component = str_c(project, component, sep = "_"))

  ),

  tar_target(obs_proj,
             obs %>%
               left_join(raw_proj, by = "project") %>%
               mutate(
                 end_inst = start_inst + time,
                 pred_na = is.na(pred)
               )
               ),


  tar_target(set_inits,
             obs_proj %>%
               mutate(
                 start = start_inst,
                 end = end_inst
               )

             ),

  tar_target(start,
             set_inits %>%
               mutate(
                 max_pred_start = map_chr(
                   pred,
                   function(p) {
                     set_inits %>%
                       filter(
                         component %in% p | pred %in% p
                       ) %>%
                       arrange(desc(end)) %>%
                       head(1) %>%
                       pull(end) %>%
                       as.character()
                   }
                 ) %>% ymd()
               )
             ),

  tar_target(
    vis,
    set_interval %>%
      ggplot() +
      geom_segment(
        aes(
          x = start,
          xend = end,
          y = component,
          yend = component
        )
      ) +
      facet_grid(project ~ .,  scales = "free")
  ),

  NULL

)