Stratified estimation of map accuracy and area
aa_card.RdStratified estimation of map accuracy and area after Card (1982). The stratified estimator accounts for disproportional allocation of samples within the strata.
Usage
aa_card(
x,
m = NULL,
w = NULL,
h = NULL,
area = NULL,
olofsson = TRUE,
se_precision = 1e-09
)Arguments
- x
Reference vector or confusion matrix if m is NULL
- m
Map vector or NULL if x is a confusion matrix
- w
Stratum weights. Named vector or specify h. If NULL, proportional sampling is assumed.
- h
Stratum names associated with stratum weights w. Alternatively, stratum names can be coded in w.
- area
total area (optional)
- olofsson
(default=TRUE) uses Olofsson formula instead of Card to estimate standard errors of reference class proportions and OA.
- se_precision
(default=1e-9) precision threshold for standard errors. Negative standard errors below that threshold are set to zero.
Value
A list of map accuracy and area proportion estimates and associated standard errors:
cm: adjusted confusion matrix in counts
cmp: adjusted confusion matrix in area proportion (sums to 1)
stats: User’s (ua) and Producer’s (pa) accuracy and the corresponding standard errors (se) for each class
accuracy: Overall accuracy and its standard error
area: estimated area proportion and standard errors for each class
References
Card, D.H. (1982). Using Known Map Category Marginal Frequencies to Improve Estimates of Thematic Map Accuracy. Photogrammetric Engineering and Remote Sensing, 48, 431-439
Olofsson, P., Foody, G.M., Stehman, S.V., & Woodcock, C.E. (2013). Making better use of accuracy data in land change studies: Estimating accuracy and area and quantifying uncertainty using stratified estimation. Remote Sensing of Environment, 129, 122-131
Examples
ex <- aa_examples("olofsson")
result <- aa_card(ex$reference, m = ex$map, w = ex$w, h = ex$h)
# or provide a confusion matrix
cm <- aa_confusion_matrix(ex$reference, ex$map)
result <- aa_card(cm, w = ex$w)