dsh-ctr-significance

In growth experiments and A/B tests, deciding whether a CTR fluctuation reflects a real lift or mere statistical noise requires a significance test. Yet that calculation often lives in scattered scripts or remote services, adding dependency overhead and friction inside an inference session. dsh-ctr-significance packages this capability as a lightweight, in-session tool so you can call it without context switches or extra deployment.

The core function, ctr_significance, performs a precise CTR statistical significance test based on the parameters you supply and returns a clearly structured, detailed result. The entire implementation is pure Node with zero network calls, so it runs reliably in offline or sandboxed environments with negligible latency.

It is well suited for growth engineers, data analysts, and strategy teams who compare landing-page variants or recommendation strategies. When you need a quick, trustworthy answer to whether an observed CTR difference is statistically significant, a single in-session call gives you the conclusion without opening a terminal or waiting on a remote service.