Topics Covered

18 practicals

Two versions of every experiment

Directory	Language	Status
this one	R — as the exam tests it	✅ Executed with R 4.3.3
python/	Python equivalents	✅ Executed and verified

Every R script is run with R 4.3.3, and the lab page shows what each printed and the plots it drew; 17_plotly.R's charts and 18_shiny_app.R are opened in Chromium by their drivers, _drive_17_plotly.py and _drive_18_shiny_app.py. Until October 2026 R could not be installed where this material is verified, and the scripts were desk-checked only. Running them found four comments whose numbers R does not give (in 04_regression.R, 10_missing_outliers.R, 13_kmeans.R and 16_arima.R); each is corrected, with a note.

What the Python equivalents buy you: the same calculation done a second way. When 02_distributions.R says a Binomial(10, 0.3) has mean 3 and variance 2.1, python/02_distributions.py computes it from the formula, and asserts it.

Run both sides yourself (it runs the R scripts too, if Rscript is installed):

pip install -r tools/data-science/requirements.txt
python3 tools/data-science/run_r_equivalents.py

Or run one R script, as the exam will, in RStudio or with Rscript 04_regression.R.

The experiments

#	Experiment	R	Python
1	Mean, median, mode, variance, SD	01_descriptive.R	python/01_descriptive.py
2	Visualise binomial, normal, Poisson	02_distributions.R	python/02_distributions.py
3	t-test and chi-square test	03_hypothesis_tests.R	python/03_hypothesis_tests.py
4	Correlation and simple linear regression	04_regression.R	python/04_regression.py
5	EDA on a real dataset	05_eda.R	python/05_eda.py
6	Feature engineering — scaling, encoding	06_feature_engineering.R	python/06_feature_engineering.py
7	Variables, control structures, functions	07_r_basics.R	— (syntax demo)
8	Read/write CSV, Excel, JSON, XML	08_file_io.R	python/08_file_io.py
9	Wrangling with dplyr and tidyr	09_wrangling.R	python/09_wrangling.py
10	Missing data and outliers	10_missing_outliers.R	python/10_missing_outliers.py
11	Dates and times	11_dates.R	python/11_dates.py
12	ggplot2 — bar, scatter, histogram, boxplot	12_ggplot.R	— (plotting)
13	K-Means clustering	13_kmeans.R	python/13_kmeans.py
14	Confusion matrix, accuracy, ROC	14_evaluation.R	python/14_evaluation.py
15	Text mining and word cloud	15_text_mining.R	python/15_text_mining.py
16	ARIMA forecasting	16_arima.R	python/16_arima.py
17	Interactive plots with plotly	17_plotly.R	— (plotting)
18	Shiny app with CSV upload	18_shiny_app.R	— (web app)

Experiments 7, 12, 17 and 18 have no Python equivalent because they demonstrate R syntax, R plotting or the Shiny framework — a translation would teach nothing.

Setting up R on your own machine

install.packages(c("tidyverse", "readxl", "jsonlite", "XML",
                   "forecast", "tseries", "plotly", "shiny",
                   "tm", "wordcloud", "caret", "pROC"))

tidyverse brings dplyr, tidyr, ggplot2, readr and lubridate in one install.