5.4-Reproducible_visualization.qmd

---
title: "Reproducible visualization"
author:
  - Elizabeth King
  - Kevin Middleton
format:
  revealjs:
    theme: [default, custom.scss]
    standalone: true
    self-contained: true
    logo: QMLS_Logo.png
    slide-number: true
    show-slide-number: all
    link-external-newwindow: true
---

## Reproducible visualization

```{r}
#| label: setup
#| echo: false
#| warning: false
#| message: false

library(tidyverse)
library(cowplot)
ggplot2::theme_set(theme_cowplot())
library(patchwork)
```

Just like your data analysis, visualization is reproducible

- Visualization is step 1 and often step *n* of analysis
- Refine exploratory plots for publication


## Git and Github

- [Happy Git and GitHub for the useR](https://happygitwithr.com/)
- [Slides for an intro to Git and Github lecture](https://github.com/egking/comp_bio_MU)
- [Video tutorial](https://youtu.be/rlXnVs1Xods)



## Your reproducible research as an R package

- Change your notion of an R package
  - Collection of code that does something
  - Functions do the analysis
  - `demo()` can run those functions
- [Package setup video tutorial](https://youtu.be/QOvZV2lkL9w)
- [Step-by-step guide](https://docs.google.com/document/d/17opEFHKGkaflOed0iww2FscIFP-_7xPKPKqcelmVAgQ/edit?usp=sharing)


## Repository structure

- Separate directories for raw data, functions, analysis, generated data, figures
- Or make a package (see above)