A comprehensive comparison of four Python packages for A/B test analysis, detailing their strengths, weaknesses, and suitability for various experimentation tasks.
About:
Evgeny Ivanov is a Principal Data Scientist with interests in optimization, marketplaces, and software engineering.
Website:
Specializations:
Interests:
Outgoing Links:
Subscribe to RSS:
The article draws an analogy between market prices and model weights in AI, emphasizing how prices can convey dispersed knowledge and solve information asymmetry in labor markets. It discusses the issue of labor-market congestion ...
0Ladder of questions in analysis
2025-05-12 •
...
0Curious facts about Ronald Fisher | Evgeny Ivanov
2024-08-12 •
The text discusses various curious facts about Ronald Fisher, including his views on smoking and lung cancer, his collaboration with William Gosset, the 'lady tasting tea' experiment, and the development of the 'tea-tasting' Pytho...
Evgeny Ivanov introduces tea-tasting, a Python package for the statistical analysis of A/B tests. The package includes statistical methods and techniques, data backends, convenient API, and detailed documentation. It provides stat...
0This meeting could have been | Evgeny Ivanov
2024-06-05 •
...
0Mediocrity and personal comfort | Evgeny Ivanov
2024-04-29 •
The text discusses the dangers of mediocrity in business and the importance of addressing it. The author shares personal experiences and reflections on the topic, emphasizing the discomfort and challenges associated with eliminati...
The blog post introduces a new Python package called tea-tasting for statistical analysis of A/B tests. It explains the motivation behind creating the package, the benefits of the Readme Driven Development (RDD) approach, and the ...
The Dunning–Kruger effect is a cognitive bias in which people with limited competence in a particular domain overestimate their abilities. The post debunks the effect by illustrating that it's a statistical artifact and the result...
The post discusses the ranking problem in Profi, a two-sided matching platform. It explains the solution to increase the number of deals in the marketplace by ranking orders based on the uplift in the probability of a deal. The po...
The text discusses the ranking problem in a two-sided matching platform, Profi, and the solution that increased the number of deals in the marketplace. It covers the concept of matching markets, the mechanics of Profi's matching p...
Rico is a Python package for creating HTML documents from rich content like dataframes, plots, images, and markdown. It provides an easy-to-use API and is useful for working with data in Python scripts. The package has no dependen...