Data

The datasets the labs run on

Textbook datasets come from the ISLR2 package in R and ISLP in Python, so you rarely need to download those. What sits here is the material specific to this course: Pakistani data, and a few classics kept locally so a lab never depends on someone else’s server being awake.

Datasets

File	Type	Size
`advertising.csv`	CSV	5 KB
`auto.csv`	CSV	18 KB
`exports-data-25-9.csv`	CSV	47 KB
`hmda-aer.dta`	DTA	858 KB
`hmda-sw.dta`	DTA	702 KB
`ict-women.dta`	DTA	3.3 MB
`kmi30-clean.xlsx`	XLSX	266 KB
`pakistan-census-complete-dataset.txt`	TXT	8 KB
`pakistan-census-separated-tables.xlsx`	XLSX	16 KB
`psx-102025.csv`	CSV	1 KB
`psx.csv`	CSV	110 KB
`table-1-national.xlsx`	XLSX	13 KB
`who.csv`	CSV	16 KB
`world-happiness.zip`	ZIP	96 KB

World Happiness panel

Annual World Happiness Report indicators, 2015 to 2024, plus a combined long file. Used in the clustering and dimension-reduction exercises.

File	Type	Size
`Readme.txt`	TXT	5 KB
`world_happiness_2015.csv`	CSV	11 KB
`world_happiness_2016.csv`	CSV	13 KB
`world_happiness_2017.csv`	CSV	13 KB
`world_happiness_2018.csv`	CSV	13 KB
`world_happiness_2019.csv`	CSV	13 KB
`world_happiness_2020.csv`	CSV	13 KB
`world_happiness_2021.csv`	CSV	12 KB
`world_happiness_2022.csv`	CSV	12 KB
`world_happiness_2023.csv`	CSV	11 KB
`world_happiness_2024.csv`	CSV	12 KB
`world_happiness_combined.csv`	CSV	134 KB

Treat data/ as read-only

Load, transform in memory, write results to outputs/. A raw file that has been edited by hand is a file whose provenance you can no longer explain, and that is the one thing a reviewer will always ask about.