Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/namakala/inappropriate-acid-suppressor-agent-use/llms.txt

Use this file to discover all available pages before exploring further.

The pipeline is declared entirely in _targets.R using the targets package. targets tracks every input and output as a node in a directed acyclic graph (DAG), so it only reruns the targets whose upstream dependencies have changed. This makes the analysis reproducible, auditable, and efficient: you can interrupt a run, change a single upstream target, and tar_make() will recompute only the affected downstream nodes.

​
Pipeline targets

The following targets are defined in _targets.R. They are listed in topological order — each target depends only on targets that appear before it.
TargetTypeDescription
fpathfilePath to data/raw/data.csv, tracked as a file dependency
tbldata frameRaw data read from fpath via readData()
tbl_cleandata frameStandardized data frame produced by clean(tbl)
mod_propmodelOverall random-effects meta-analysis of proportions
mod_copas_propmodelCopas selection model applied to mod_prop
plt_forest_propplotForest plot saved to docs/figures/meta-analysis-prevalence.pdf
plt_funnel_propplotFunnel plot for the overall model
mod_subgroup_*modelSubgroup meta-analysis for each variable in uni_vars
mod_copas_subgroup_*modelCopas-adjusted model for each subgroup
plt_forest_subgroup_*plotForest plot for each subgroup, saved to docs/figures/
plt_funnel_subgroup_*plotFunnel plot for each subgroup
mod_metareg_*modelUnivariable meta-regression for each variable in uni_vars
mod_metareg_mvmodelMultivariable meta-regression using all variables in mv_vars
reportdocumentCompiled Quarto PDF report from docs/report.qmd
readmedocumentRendered README.qmd (lowest priority, runs last)
The subgroup and univariable meta-regression targets are generated dynamically using tar_map() over the uni_vars vector: 
uni_vars <- c(
  "Age", "Continent", "Setting", "JBI_Classification", "use_guideline"
)
The multivariable meta-regression uses a wider set of predictors: 
mv_vars <- c(
  "Year", "JBI_Classification", "use_guideline", "Setting", "Continent",
  "Sample_size"
)

​
Pipeline DAG

The diagram below shows how targets flow from raw data to final outputs.

​
Pipeline configuration

The full tar_option_set() block from _targets.R: 
tar_option_set(
  packages   = pkgs,
  error      = "continue",
  memory     = "transient",
  controller = crew_controller_local(worker = 4),
  storage    = "worker",
  retrieval  = "worker",
  garbage_collection = TRUE
)
Each option controls a specific aspect of pipeline behavior:
OptionValueEffect
packagesmagrittr, targets, tarchetypes, crewLoaded in every worker before a target runs
error"continue"A failing target marks itself as errored but does not halt the rest of the pipeline
memory"transient"Target objects are removed from memory after they are no longer needed by downstream targets
controllercrew_controller_local(worker = 4)Spawns up to 4 parallel local R worker processes
storage"worker"Workers write results directly to the _targets/ store without routing through the main process
retrieval"worker"Workers read upstream dependencies directly from the store
garbage_collectionTRUECalls gc() after each target to release memory promptly

​
Parallel execution

The pipeline uses crew to manage parallel workers. The controller is configured with: 
crew_controller_local(worker = 4)
This spawns up to four background R processes on the local machine. targets dispatches independent targets — such as the five mod_subgroup_* targets or the five mod_metareg_* targets — to available workers concurrently. Because storage = "worker" and retrieval = "worker" are both set, workers communicate directly with the _targets/ object store rather than sending large data objects through the main R session. This avoids the serialization bottleneck that would otherwise occur when transferring model objects between processes.
The memory = "transient" and garbage_collection = TRUE settings work together with parallel execution to keep per-worker memory usage low when fitting many subgroup models simultaneously.

​
Next steps

Data preparation

Learn how to format and place your input CSV, and how the clean() function standardizes variables before modeling.

Running the workflow

Commands for running, inspecting, and troubleshooting the targets pipeline.

Build docs developers (and LLMs) love

Get started for freeTalk to us