>>>>>>>> SAS Notes -- Copyright (c) 1991 - 1999 by SAS Institute Inc. <<<<<<<<
V7-ARIMA-0485 *** ALERT NOTE *** PRODUCT: ETS
TITLE: Error or incorrect forecasts when input series has missing values at end
KEYS: ERROR INCORRECT FORECASTS MISSING VALUES INCORRSTAT INCORROUT INVALID
When forecasting a transfer function model in which the input series
has missing values at the end of the data, but the response series does
not, PROC ARIMA may return an Invalid Operation error, a Domain error or
incorrect forecasts for the response series.
For example, suppose a data set has 46 observations in total. Y, the
response series, has complete data for all 46 observations. X, the
input series, has complete data through observation 40, but missing
values for observations 41-46. If you try to forecast Y through
observation 50, PROC ARIMA will result in an Invalid Operation error, a
Domain error or in incorrect forecasts for Y.
To circumvent the problem, identify, estimate and forecast a model for
X, merge the forecasts for X with the original data, and then estimate
and forecast the model for Y. Note that the standard errors of the
forecasts for Y will be underestimated.
SYSTEM RELEASE LEVEL RELEASE LEVEL
AIX/R 7.00 0B1 .
CMS 7.00 0B1 .
DigitUnx 7.00 0B1 .
HP800 7.00 0B1 .
MVS 7.00 0B1 .
OS/2 7.00 0B1 .
Solaris 7.00 0B1 .
VMS/AXP 7.00 0B1 .
VMS 7.00 0B1 .
WIN/NT 7.00 0B1 .
Win98 7.00 0P1 .
Win95 7.00 0B1 .
This archive was generated by hypermail 2b29 : Mon Jan 10 2000 - 10:10:19 EST