View Code? Open in Web Editor
NEW
A/B testing for Human
Home Page: https://farhanreynaldo.github.io/abby/
License: MIT License
Makefile 4.50%
Python 95.36%
Shell 0.15%
abby's Introduction
Hi there ๐ I'm Farhan Reynaldo Hutabarat
๐ซ I'm currently pursuing Masters in Quantitative Methods in the Social Sciences at Columbia University.
๐ผ Previously, I worked as a senior data analyst in Tokopedia and as a data scientist in Delman Data Lab and Dattabot .
๐ฑ My interest lies in the field of bayesian data analysis, experimentation, and causal inference.
๐ซ You can get in touch with me by email at [email protected] and LinkedIn .
abby's People
Contributors
Watchers
abby's Issues
The argument for metrics across the compare_* methods is inconsistent. The compare_ttest
have the numerator
argument, which is confusing since it mimics the naming convention for compare_delta
. There should be more intuitive naming for the metrics that work across all compare_* methods.
Every compare_*
method should have a unified statistics result, which consists of:
Control mean
Experiment mean
Control standard deviation
Experiment standard deviation
Diff between means
Diff means 95% confidence interval
P-value
Add example for A/B/N testing in the README file.