day4-analytical_workflows.qmd

---
title: "Reproducible Analytical Workflows"
---


## Leveraging existing frameworks for reproducibility

Please remember that every workflow or framework is opinionated. This is also true for the tools out there that can help you to develop reproducible workflows. It is great practice to test them on a small project to see if hey could fit your project needs and way of working and collaborating. In this section we will go over a few R packages that could be of interest for your project. 


### Analytical workflow frameworks in R

```{r tidy-workflow, out.width='80%', fig.align="center", fig.cap="Conceptual workflow model promoted by the `tidyverse`. Note the program box around the workflow and the iterative nature of the analytical process described. _Source: R for Data Science <https://r4ds.hadley.nz/intro#fig-ds-diagram>_",echo=FALSE}

knitr::include_graphics("img/tidy-workflow.png")
```


#### Targets

The `targets` R package can be very useful if you have a complex workflow that is built of many parts that take time to rerun. Target can detect and run only the strict necessary steps to rerun when a specific change has been done to the workflow.

Here for more: <https://books.ropensci.org/targets/>

**Our practice example: <https://github.com/EDS-214/eds214-handson-targets>**


#### Pointblank

Although the `pointblank` package is mainly framed as a validating tool of the various part of your workflow, it provide a set of tools and great integration with the R Markdown ecosystem. It is also quite flexible making it possible to leverage this tool in a variety of project setups.

<https://github.com/rstudio/pointblank>