Computes predictions with confidence and prediction intervals.
The intervals are computed as described in the module documentation using the leverage of each prediction point and the critical value of Student's t distribution. The covariance matrix and the residual mean squared error are taken from the model if it was created by doe_fit_model; otherwise, x_fit and residual_mse must be supplied. If either is supplied, it takes precedence over the value stored in the model. The routine halts with an error if the required information is not available, or if the residual degrees of freedom are not positive.
| Type | Intent | Optional | Attributes | Name | ||
|---|---|---|---|---|---|---|
| class(doe_model), | intent(in) | :: | mdl |
The DOE model. |
||
| real(kind=real64), | intent(in), | dimension(:,:) | :: | x |
The M-by-N matrix containing the M values of each of the N factors at which to evaluate the model. |
|
| real(kind=real64), | intent(in), | optional | :: | alpha |
The significance level, in the range (0, 1). The default is 0.05 (95% intervals). |
|
| real(kind=real64), | intent(in), | optional, | dimension(:,:) | :: | x_fit |
An optional K-by-N matrix containing the factor values used to fit the model. This is used to compute the parameter covariance and the residual degrees of freedom . |
| real(kind=real64), | intent(in), | optional | :: | residual_mse |
An optional residual mean squared error of the fit. |
A doe_prediction object containing the M predicted values, the lower and upper confidence bounds, the lower and upper prediction bounds, and the confidence level.