interpolationErrorFunctions: Compare original and interpolated maps point-wise

In sensors4plumes: Test and Optimise Sampling Designs Based on Plume Simulations

Description

absError and delineationError compare an interpolated to an original map (the values in each point respectively). They are to be used as fun_error in interpolationError. absError is the absolute difference between the values. delineationError compares areas above a given threshold, and indicates false positive and false negative classification.

absErrorMap and delineationErrorMap are to be used as fun_l in interpolationError, they compute the average over all plumes of the respective error functions to generate a common map.

Usage

absError(x, nout = 1)
delineationError(x, nout = 1, threshold = 1e-7, weightFalseNeg = 5)
absErrorMap(x, nout = 1)
delineationErrorMap(x, nout = 1, threshold = 1e-7, weightFalseNeg = 5)

Arguments

x	vector of length 2 with x[1] the value and x[2] the interpolated value
nout	length of output – needed as the function is to be called via simulationsApply as fun_pl or fun_l
threshold	threshold to classify all locations
weightFalseNeg	weight for false negative classification, false positive classification is weighted 1

Value

absError returns a single numeric value, the absolute difference.

delineationError first computes if the original is above the threshold and if the interpolated is above the threshold; then it determines the result: 0 if classifications agree, 1 for false positive (i.e. original is below threshold, interpolated is above), and weightFalseNeg for false negative.

absErrorMap and delineationErrorMap return single values.

Author(s)

Kristina B. Helle, [e-mail]

Examples

## see interpolationError

sensors4plumes documentation built on May 1, 2019, 10:27 p.m.