spike_median: Modified Hampel Median Filter.

In SandroGroth/phenoRS: Preprocessing and Extraction of phenological parameters from Remote Sensing Data.

Description

Modified Hampel Median Filter.

Usage

spike_median(v, w, ypts, w_min = 0, spk = 2, cpp = T)

Arguments

v	numeric vector. Vegetation index values.
w	numeric vector. Initial weights.
ypts	numeric. The number of values per year. Used to determine the moving window size.
w_min	numeric. Specifies the weight that should be assigned to outliers.
spk	numeric. The Spike parameter. the higher, the more values are classified outliers.

Details

#TODO

Author(s)

Sandro Groth

References

Hampel F. R., ”The influence curve and its role in robust estimation,” Journal of the American Statistical Association, 69, 382–393, 1974

Eklundh, L., and Jönsson, P., 2017, TIMESAT 3.3 with seasonal trend decomposition and parallel processing - Software Manual. Lund University, 92 pp.

Examples

## Create a random normal distribution with outliers.
values <- c(rnorm(19), 50, rnorm(19), -50, rnorm(20))

## set all initial weights to 1
init_weights <- rep(1, 60)

## get the spike filtered weights
filtered_weights <- spike_median(values, init_weights, 20, 0, 2)

SandroGroth/phenoRS documentation built on July 12, 2020, 4:33 a.m.