Experimental Design (DOE & randomization)

experimental-design — Experimental Design (DOE & randomization)

Generate seeded randomization/allocation schedules and design-of-experiments matrices (read-only, local). Each perk reads a small JSON spec and writes a reproducible CSV layout under record_store.

What to look out for

Each tool emits one line of structured JSON (the audit + debug log) and writes its artifacts under record_store. LOGS TO CHECK: that line + the named CSV + the executor run-ledger. Every layout is seeded, so the exact schedule/matrix can be archived and regenerated (a requirement for trial registration and good lab practice).

The randomization perks (simple-, block-, stratified-block-, cluster-randomization, assign-factorial-runs) need only numpy + pandas. The DOE-matrix perks (full-, two-level-, fractional-factorial, plackett-burman, central-composite, box-behnken, latin-hypercube) additionally require pyDOE3; when it is absent the porter writes a graceful empty {} artifact rather than failing.

Perks

perk	tool	nature
simple-randomization	simple_randomization	read-only — independent per-unit assignment (numpy/pandas)
block-randomization	block_randomization	read-only — permuted blocks, balance throughout (numpy/pandas)
stratified-block-randomization	stratified_block_randomization	read-only — blocks within strata (numpy/pandas)
cluster-randomization	cluster_randomization	read-only — randomize whole clusters (numpy/pandas)
assign-factorial-runs	assign_factorial_runs	read-only — randomize run order of design rows (numpy/pandas)
full-factorial	full_factorial	read-only — full factorial over explicit levels (pyDOE3)
two-level-factorial	two_level_factorial	read-only — full 2^k factorial (pyDOE3)
fractional-factorial	fractional_factorial	read-only — 2^(k-p) from a generator (pyDOE3)
plackett-burman	plackett_burman	read-only — main-effects screening (pyDOE3)
central-composite	central_composite	read-only — response-surface CCD (pyDOE3)
box-behnken	box_behnken	read-only — response-surface BBD, >=3 factors (pyDOE3)
latin-hypercube	latin_hypercube	read-only — space-filling sample (pyDOE3)

All perks are destructive: false: they read a JSON spec and write a CSV; nothing mutates a remote or live service.

How to use it

Pick a perk, copy ledger.json → task-ledger.json, fill its vars + record_store, then validate → compose → compile → oversight → executor.

Localized from K-Dense-AI/scientific-agent-skills experimental-design — MIT (see LICENSE.txt).