Loading

RLOWESS

gts

The RLOWESS function applies an iterative smoothing program on a Geo Time Series™ or a LIST of Geo Time Series™, which is done via locally weighted regression, and is robust to outliers.

REFERENCE

Cleveland, W. S. (1979) Robust locally weighted regression and smoothing scatter plots. J. American Statistical Association 74, 829–836.

Cleveland, W. S. (1981) LOWESS: A program for smoothing scatter plots by robust locally weighted regression. The American Statistician 35, 54.

RLOWESS is available since version 1.0.0.

See also

Signatures

Examples

// a linear GTS NEWGTS 1 50 <% NaN NaN NaN 4 PICK TODOUBLE ADDVALUE %> FOR // we add outliers CLONE 12 NaN NaN NaN 0 ADDVALUE 24 NaN NaN NaN 100 ADDVALUE 35 NaN NaN NaN 500 ADDVALUE DEDUP SORT 35 30 0 1 RLOWESS

Examples

// a linear GTS GTS = NEWGTS() FOR(1, 50, (i) -> { ADDVALUE(GTS, i, NaN, NaN, NaN, TODOUBLE(i)) } ) // we add outliers (original, clone) = CLONE(GTS) ADDVALUE(clone, 12, NaN, NaN, NaN, 0) ADDVALUE(clone, 24, NaN, NaN, NaN, 100) ADDVALUE(clone, 35, NaN, NaN, NaN, 500) clone = DEDUP(clone) clone = SORT(clone) return original, RLOWESS(clone, 35, 30, 0, 1)