bias.all {QFASA} | R Documentation |
beta.meths.CI
.Calculate bias correction for confidence intervals from beta.meths.CI
.
bias.all(p.mat, prey.mat, cal.mat = rep(1, length(ext.fa)), fat.cont = rep(1, nrow(prey.mat)), R.bias, noise, nprey, specify.noise, dist.meas, ext.fa)
p.mat |
matrix containing the fatty acid signatures of the predators. |
prey.mat |
matrix containing a representative fatty acid signature |
cal.mat |
matrix of calibration factors where the i th column is to be used with the i th predator. If modelling is to be done without calibration coefficients, simply pass a vector or matrix of ones. |
fat.cont |
prey fat content |
R.bias |
botstrap replicates |
noise |
noise |
nprey |
number of prey |
specify.noise |
noise |
dist.meas |
distance measure |
ext.fa |
subset of FA's to use. |
Row 1 is Lambda CI, row 2 is Lambda skew, and row 3 is Beta CI
## Fatty Acids data(FAset) fa.set = as.vector(unlist(FAset)) ## Predators data(predatorFAs) tombstone.info = predatorFAs[,1:4] predator.matrix = predatorFAs[, fa.set] npredators = nrow(predator.matrix) ## Prey prey.sub = preyFAs[, fa.set] prey.sub = prey.sub / apply(prey.sub, 1, sum) group = as.vector(preyFAs$Species) prey.matrix.full = cbind(group,prey.sub) prey.matrix = MEANmeth(prey.matrix.full) ## Calibration Coefficients data(CC) cal.vec = CC[CC$FA %in% fa.set, 2] cal.mat = replicate(npredators, cal.vec) # Note: uncomment examples to run. CRAN tests fail because execution time > 5 seconds # diet.est <- p.QFASA(predator.mat = predator.matrix, # prey.mat = prey.matrix, # cal.mat = cal.mat, # dist.meas = 2, # start.val = rep(1,nrow(prey.matrix)), # ext.fa = fa.set)[['Diet Estimates']] # # bias <- bias.all(p.mat = diet.est, # prey.mat = prey.matrix.full, # cal.mat = cal.mat, # R.bias = 10, # noise = 0, # nprey = 50, # dist.meas = 2, # ext.fa = fa.set)