Readme file
SERIES C
 Applied Statistics
Estimation and prediction for low degree polynomial models under
measurement errors with an application to forest harvesters
T. Nummi and J. Möttönen
Appl. Statist., 53 (2004) 495 - 505
The programs are divided into two sets: the first set of programs are
associated with Table 1 in the paper and the second set is associated
with Table 2.
Table1
======
There are tree data sets in three folders:
1) manu (manually measured stem data)
2) h1 (measured by harvester 1)
3) h2 (measured by harvester 2)
SAS programs are given in the folder manu. To execute these files in SAS
appropriate paths of files to be used should be given in program calls.
Table2
======
The second set of programs (Table2) consists of a macro (file: macro_final),
which is used to generate the coefficient folders 1a, 1b, 2a and 2b. The macro
name is exclude and it should be submitted first (note the correct paths of
the associated files in exclude). Thereafter submit the macro by submitting
%exclude(26,'file'). Now the resulting file contains the model parameter estimates
when one stem in turn is excluded.
The R-programs are used with two arguments. The first argument is the name of
the file (stem file) and the second argument is the file of estimated model
coefficients (when 1 stem is left out in turn).
Tapio Nummi
Tampere School of Public Health
University of Tampere
Tampere
FIN-33014
Finland
E-mail: tapio.nummi@uta.fi
|
