MB-datasets

Small datasets for teaching and training at Melbourne Bioinformatics, The University of Melbourne.

Purpose

Curated, ready-to-use datasets for workshops, tutorials, and course materials. Each dataset is small enough to be bundled directly in a repo (under 1 MB) and comes with provenance documentation.

Datasets

Dataset	Description	Rows	Format
palmerpenguins	Morphological and isotope measurements for three Antarctic penguin species	344	CSV

Downloading a dataset

All files are accessible directly via raw.githubusercontent.com. The URL pattern is:

https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/<dataset-name>/<filename>

wget

wget https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/palmerpenguins/penguins_lter.csv

curl

curl -O https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/palmerpenguins/penguins_lter.csv

Python

import pandas as pd
url = "https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/palmerpenguins/penguins_lter.csv"
df = pd.read_csv(url)

R

url <- "https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/palmerpenguins/penguins_lter.csv"
df <- read.csv(url)

For gzipped files (.csv.gz), both wget/curl download as-is; pandas and R's read.csv decompress automatically.

---

Contributing

1. Create a new directory under datasets/ using a short, lowercase, hyphenated name (e.g. datasets/my-dataset/).
2. Add a README.md — see datasets/palmerpenguins/README.md for the expected format. At minimum it must include:
   • A URL to the original source (ideally one with full metadata/provenance)
   • A table of columns with descriptions
   • License information
3. Keep dataset files under 1 MB. Gzip (.gz) is fine if needed.
4. Add a row to the table above.

README template

# Data: <Dataset Name>

## File

| File | Rows | Description |
|------|------|-------------|
| `data.csv` | N | Brief description |

## Download

```bash
# wget
wget https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/<dataset-name>/<filename>

# curl
curl -O https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/<dataset-name>/<filename>

# Python / pandas
import pandas as pd
df = pd.read_csv("https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/<dataset-name>/<filename>")

# R
df <- read.csv("https://raw.githubusercontent.com/melbournebioinformatics/MB-datasets/main/datasets/<dataset-name>/<filename>")

Provenance

Downloaded from:

Original source: <institution / author / study>

Citation:

Author et al. (Year). Title. Journal vol(issue): pages. https://doi.org/...

License

Columns

Column	Description
col_name	What it contains

Why this dataset?

<One paragraph: what makes it useful for teaching?>

## License

Individual datasets carry their own licenses — see the `README.md` in each subdirectory. Repository infrastructure (this file, templates) is released under [CC0](https://creativecommons.org/publicdomain/zero/1.0/).