Planned Experiment

PlanExp is a complete environment for design of experiments, ordinary least squares regression, and response-surface work — from the planning matrix to 2D contours, 3D surfaces, and ternary diagrams. Build a plan, enter replicate measurements, fit a polynomial model, test it, and read the optimum inside the coded cube [−1, +1]^k. Factor coding follows the classical DOE convention x = (X − X0) / ΔX. WHAT YOU CAN DO • Choose a design: one-factor grid, 2^k and 2^(k−p), Box–Draper B-D13, rotatable / face-centreed / inscribed CCD, Box–Behnken, Plackett–Burman, Doehlert, 3^k, Taguchi arrays, simplex lattice for mixtures, D-optimal, Latin hypercube, and definitive screening. • Fit OLS models with main effects, two-factor interactions, and quadratic terms. Optional three-way interactions for 2^k. • Run Cochran’s C for variance homogeneity, Student’s t on coefficients, and Fisher’s F for lack of fit. See R², adjusted R², and RMSE. • Drop insignificant terms and refit. Canonical analysis locates the stationary point; a constrained search reports the best feasible point in the experimental region. • Plot response curves, filled contours, interactive 3D surfaces, and ternary (mixture) diagrams in natural units. FREE AND FULL The free version includes a one-factor grid, full factorial 2^k with up to three factors, two parallel runs, PDF reports, and JSON of your own project. Local gallery save is always free. The full version unlocks every design (including a custom matrix), up to 12 parallel runs, and CSV export — a single lifetime purchase. WHO IT IS FOR Researchers, process engineers, chemists, metallurgists, and students who plan laboratory or plant trials and need transparent classical statistics rather than a black-box optimiser. PlanExp continues the lineage of the PlanExp 2³ and PlanExp B-D13 desktop tools, rewritten for iPhone, iPad, and Mac with a shared math core.