DatSciTrain_Rtargets_prep_tool

Anh Han

---

Introduction

The Rtargets_prep_tool is a lightweight template for starting small R projects. It uses an Excel-based planning interface to help structure workflows and streamline project setup with the targets package.

This tool was developed to:

• Support researchers and analysts in planning data pipelines before coding;

• Simplify management of complex data analysis workflows in R;

• Automate the generation of targets blocks and utility function shells;

• Promote good data management practice;

• Integrate with version control for track changes and collaborative development.

Contents

The repository contains the following core components:

Rtargets_prep_tool/
├── Rtargets_prep_tool.xlsx      # Excel spreadsheet for defining targets, inputs, and utilities
├── _targets.R                   # Main pipeline file (insert code snippets here)
├── config.R                     # Centralised configuration (e.g., file paths, global settings)
├── run.R                        # Script for running, debugging, and inspecting the pipeline
├── R/                           # Folder where utility functions are generated
├── load_packages.R              # Script for loading required R packages

How to use

1. Download or clone this repository by clicking the green <> Code button at the top right of this Github repository file listing. You will need to create your R working folder and rename the folder and .Rproj file as needed.

2. Launch Rtergats_prep_tool.xlsx and enter each pipeline step in the yellow-highlighted rows. Use the verb_noun naming convention (e.g., load_ap_monitors) and specify any required input files or upstream targets.

Column D: Generates target code snippets for _targets.R

Column E: Creates utility functions in the R/ folder 

1. Paste the content from Column D into _targets.R, inside the list(...) block. Then copy and run the code from Column E in the R console to generate .R files inside the R/ folder. Each .R file corresponds to one pipeline step and will contain a scaffolded function ready for you to fill in.

2. To implement the logic inside each function, open each generated .R file in the R/ folder and fill in your data processing or analysis logic. Remember to use proper return() statements to return objects, passing through the pipe.

NOTE: if you are woking with spatial objects (like terra and sf), wrap the object with return(wrap(spatial_object)) and unwrap in the downstream function: spatial_object <- unwrap(name_of_previous_target).

1. Run and inspect the pipeline by using run.R as your main control script. It typically includes source(run.R). Inside run.R, you can run the pipeline with tar_make(), load targets with tar_load("target_name"), and visualise dependencies with tar_visnetwork().

Remember to update config.R with file paths or constants. You can also use load_packages.R to load required libraries in interactive sessions.

1. Revisit the Excel file to add or modify targets and use Git to track and version updates made to _targets.R, R/, and config.R.

Example application: GIS analysis of air pollution monitor coverage

An example of this tool is demonstrated in the project DatSci_ap_monitors_and_postcodes, which use buffer analysis around monitoring stations and spatial joins with postcode boundaries.

Training sessions applying the Rtarget_prep_tool for this analysis are documented in the transcript.

This site is open source. Improve this page.