ergodicDocs
Products: switch from ergodic to another product
Get started

NotebooksNotebooks

Notebooks

Download a notebook and run it yourself. Each preview shows the cells with outputs cleared; release CI runs every notebook.

Learn series#

Why a correlation can't answer your question#

Good and bad controls#

Estimating effects: from a graph to a number#

Who responds? Heterogeneous effects#

Hidden confounding: when the back door has no handle#

Discovering structure: when nobody can draw the graph#

Quasi-experiments: when the rollout is the experiment#

Process mining: your operation in three steps#

Process intelligence: when the process is the question#

Decision points: when the process picks a lane#

Forecasting: a number for next quarter, and where it came from#

Structural causal models: one fit, three kinds of question#

Hidden regimes: a switch nobody logged#

Causal forests and policy trees: from per-customer effects to a rule#

Case studies#

Does smoking lower birthweight? A back-door adjustment, start to finish#

Berkeley, 1973: the admissions paradox that wasn't#

Does a 401(k) make you richer? The pension puzzle#

Proposition 99: building a California that never passed the law#

Discovery on a known network: grading the graph#

Customer lifetime value: spend where it moves#

Customer churn: when the fix looks like the cause#

Suggest an edit
Documentation

Search ergodic 0.2.1

Search ergodic by task, function or error.