Skip to content

API Reference

fastlowess-wasm


fastlowess-wasm / OnlineSmoothOptions

Defined in: fastlowess_wasm.d.ts:134

Configuration options for online LOWESS smoothing. A subset of SmoothOptions: diagnostics, residuals, parallel execution, cross-validation, and the sorted output have no equivalent here. The se output and intervals require update_mode = "full".

optional auto_converge?: number

Defined in: fastlowess_wasm.d.ts:152

Auto-convergence tolerance. Disabled when absent.


optional boundary_policy?: string

Defined in: fastlowess_wasm.d.ts:148

Boundary handling (“extend”, “reflect”, “zero”, “noboundary”). Default: “extend”.


optional delta?: number

Defined in: fastlowess_wasm.d.ts:140

Delta for interpolation speedup. Default: auto. Set to 0 to disable.


optional fraction?: number

Defined in: fastlowess_wasm.d.ts:136

Smoothing fraction (0 < fraction <= 1). Default: 0.67.


optional intervals?: IntervalsOptions

Defined in: fastlowess_wasm.d.ts:156

Confidence/prediction levels and per-window residual-bootstrap refits. Requires update_mode = "full".


optional iterations?: number

Defined in: fastlowess_wasm.d.ts:138

Number of robustness iterations. Requires update_mode = "full". Default: 0.


optional missing?: string

Defined in: fastlowess_wasm.d.ts:160

Policy for non-finite (NaN/Inf) x/y values passed to add_point (“error”, “drop”). Default: “error”.


optional outputs?: string[]

Defined in: fastlowess_wasm.d.ts:154

Optional output components: se, weights, derivative.


optional robustness_method?: string

Defined in: fastlowess_wasm.d.ts:144

Robustness method (“bisquare”, “huber”, “talwar”). Default: “bisquare”.


optional scaling_method?: string

Defined in: fastlowess_wasm.d.ts:150

Scaling method (“mad”, “mar”, “mean”). Default: “mad”.


optional seed?: number | bigint

Defined in: fastlowess_wasm.d.ts:158

Bootstrap safe-integer Number or unsigned 64-bit BigInt seed; each full-update window restarts from it.


optional weight_function?: string

Defined in: fastlowess_wasm.d.ts:142

Kernel function (“tricube”, “epanechnikov”, “gaussian”, “uniform”, “biweight”, “triangle”, “cosine”). Default: “tricube”.


optional zero_weight_fallback?: string

Defined in: fastlowess_wasm.d.ts:146

Fallback when all weights are zero (“use_local_mean”, “return_original”, “return_none”). Default: “use_local_mean”.