Skip to content

Best Practices

There are a variety of hardware and software options and combinations available to you. While there are few concrete rules, here is an attempt to guide your success.

Resources for thinking about open and reproducible scientific computing.

Practical Aspects

Foremost, code should be written so someone else could easily run. This means they have access and can understand.

1) Code should be in a Github repository

2) Organize your data and code.

Here's an example of how repos should be organized. Note that each directory contains a README.md, which describes the contents of each directory (and, sometimes even describes each file in that directory).

Create a directory called gitrepos and then keep all subsequent repositories within it.

gitrepos$ tree

.
├── <GitHub username>
│   └── Project-01
│       ├── code
│       │   ├── 01-FastQ-QC.Rmd
│       │   ├── 02-DESeq2
│       │   └── README.md
│       ├── data
│       │   ├── genome-features
│       │   │   └── README.md
│       │   ├── raw
│       │   │   └── README.md
│       │   └── README.md
│       ├── outputs
│       │   ├── 01-FastQ-QC
│       │   │   ├── figures
│       │   │   │   └── README.md
│       │   │   └── README.md
│       │   ├── 02-DESeq2
│       │   │   ├── figures
│       │   │   │   └── README.md
│       │   │   └── README.md
│       │   └── README.md
│       └── README.md
└── RobertsLab
    ├── lab-website
    │   └── README.md
    ├── README.md
    ├── resources
    │   └── README.md
    └── tusk
        └── README.md

Working with Git and RStudio on Raven

related - getting a Personal Access token - https://d.pr/i/lS8UAr

Papers

Workshops

Courses