Bootstrapping is an intuitive and formula-free way of calculating confidence intervals, hypothesis testing and sample sizes without the need for large samples or Normal distributions.

How stratified sampling and cluster sampling works, with examples.

Information on water points is available on Kenya's open data portal. The functional status of water points can be functional, not functional or something in between (dodgy). A classifier was built that labels the functional status of water points. The accuracy of this classifier is 92 percent, the overall precision 88 percent.

Test parcoords library with different layouts

Dashboard using data from UNEP/RISOE and IGES.

Time series analysis of rainfall per dekad (10 days) over 30 years. There is some evidence that the rainy season is starting to arrive later than it used to.

How to set up R, install packages and work with data.table.

Sample data is analyzed with OLS. Diagnostic plots & remediation of non-conformities with OLS assumptions.

Combine data from FAO, World Bank and the Government of Kenya to generate an interactive report.

Reporting in most organizations involves a database/Excel combination. This process is resource-intensive, error-prone and it outputs boring reports. It also degrades trained people to data monkeys. There is good news, however. The entire pipeline, including quality control can be done in R and made interactive. This demo shows how.

Presentation for Upande.com in Nairobi

Plotting a large series of objects on a world map defines the world map.

Sets out the basics of data.table.

Overview of operations in data.table. The standard documentation in data.table is excellent, but here data.table is applied on a real world dataset: 337 K flights from NYC.