Can I ask another SAS question.
A common technique is to augment the data matrix with extra rows of values
for which prediction is required. The program (PROC GLM) removes those
rows with missing y-values before doing the sums but still calculates
fitted values and CLI confidence intervals.
But where there are weights involved, how are these handled? Normally the
weights will add up to the number of rows of the data matrix before
augmentation. But will SAS instead include the weights of the additional
rows when normalising the weights? Or will it (correctly) ignore them?
And when estimating a confidence interval for these augmented rows, what
will happen? Because normally you would not use weights when predicting
values - the weights are there because you believe some rows are more
important than others (eg a lower variance) and affect the calibration of
the model.
Or should the ESTIMATE command be used instead.
Can anyone point me to the appropriate part of the SAS manual that
describes what actually happens?
TIA
John
John Logsdon "Try to make things as simple
Quantex Research Ltd, Manchester UK as possible but not simpler"
[log in to unmask] [log in to unmask]
+44(0)161 445 4951/G:+44(0)7717758675 www.quantex-research.com
|