adam::blog

He was only five years old when the plague broke out so could not remember it. To describe it in his novel, nearly fifty years later, Daniel Defoe needed to dust off faded diaries, survey old medical pamphlets and query archival statistics. Well done research paid off. For decades the readers of "A Journal of the Plague Year" were thinking they are reading genuine memoirs of the 1665`s Great Plague eyewitness. However, concerning statistics included in the "Journal", they cannot be said deluded. These were real and for their substantial use, we can call the "Journal" a data-based fiction. And data-based fiction is nev er closed for augmentation: data can speak more than the author transmitted. I want to give them a voice once again in this letter. A letter to "Journal`s" narrator, from a data scientist writing in the second plague year. [Read More]

In this post I take a look at the resolution providing free transport for car drivers in case of huge air pollution, adopted in Krakow in 2015. Three versions of this resolution have been in use. In the first version, decision about providing free transport was on a basis of measurements from the previous day. But this method was criticized for having one-day lags in decisions. In the next version decisions were based on forecasts. But what was criticized then was the accuracy of those forecasts. From 5th July 2018 both criteria are used. Whih is a good idea as I show below. [Read More]

When I was learning how to work with APIs in R few months ago, I discovered there is an API for tram locations and departures in my home town - Kraków. Since then I’ve been planning the analysis of tram delays based on it and eventually I’ve taken it as my personal summer project. The entire report in Polish is dozens page long. Here, on the English side of the blog, I omit the overview of data preparation and technical details, which are important but not so much interesting and present several selected insights. [Read More]