User`s guide

predint
4-123
4predint
Purpose Compute prediction bounds for new observations or for the function
Syntax ci = predint(fresult,x)
ci = predint(fresult,x,level)
ci = predint(fresult,x,level,'intopt','simopt')
[ci,ypred] = predint(...)
Arguments
Description
ci = predint(fresult,x) returns prediction bounds for new response
(observation) values at the predictor values specified by
x. The confidence level
of the predictions is 95%.
ci contains the upper and lower prediction bounds.
fresult is the fit result object returned by the fit function. You can compute
prediction bounds only for parametric fits. To compute confidence bounds for
the fitted parameters, use the
confint function.
ci = predint(fresult,x,level) returns prediction bounds with a confidence
level specified by
level.
ci = predint(fitresult,x,level,'intopt','simopt') specifies the type of
bounds to compute. If
intopt is functional, the bounds measure the
uncertainty in estimating the function (the fitted curve). If
intopt is
observation, the bounds are wider to represent the additional uncertainty in
predicting a new response value (the fitted curve plus random noise).
fresult A fit result object.
x The values at which predictions are calculated.
level Confidence level. The value must be between 0 and 1. The
default value is 0.95.
'intopt' Can be observation (the default) to compute bounds for new
response values, or
functional to compute bounds for the fit
evaluated at
x.
'simopt' Can be off (the default) to compute nonsimultaneous
bounds, or
on to compute simultaneous bounds.
ci An array of upper and lower prediction bounds.
ypred The predicted (fitted) value of fresult evaluated at x.