Readme file
SERIES C
 Applied Statistics
The analysis of designed experiments and longitudinal data by using
smoothing splines (with discussion), by A. P. Verbyla, B. R. Cullis, M. G. Kenward
and S. J. Welham
Journal of the Royal Statistical Society, Series C, Applied Statistics,
volume 58 (1999), part 3, pages 269-311
There are three data sets called:
cow.dat, ELISA.dat and cowling.dat.
Arunas P. Verbyla
BiometricsSA
University of Adelaide and SARDI
Private Mail Bag 1
Glen Osmond
SA 5064
Australia
E-mail: averbyla@BiometricsSA.adelaide.edu.au
Data on dairy cows: The data file contains 11 columns (variables) on each line, with
the top row labelling the variables, namely
y liveweights (in kgs)
trt the 2x2 factorial treatments, 1,2,3,4
time time as a factor with 23 levels
animal cow as a factor with 26 levels
ctime (actual time-122)/10
x centred (zero mean) and scaled (divided by 100) times
infect infection factor with 2 levels, text
iron iron factor with 2 levels, text
time3 actual time - 155, with first two values set to zero
i3 indicator for iron application (0 for times 1 and 2, 1 otherwise)
rtime3 as for time3 BUT with NA or missing for times 1 and 2
There are 598 lines of data and NA stand for missing data.
Multi-environment lupins/sedding rate trial: There are 12 columns (variables) per line
in the data file, with the top row labelling each column as follows.
row factor indicating the row position in the field trial within site
column factor indicating the column position in the field trial within site
variety variety as a factor (alpha-numeric)
seedrate seeding rate
lincol centred column (mean 0) position as a variate
linrow centred row (mean 0) position as a variate
rancol column as a factor (random factor); identical to column
ranrow row as a factor (random factor); identical to row
fixedrow1 factor of 4 levels for serpentine seeding effect
x (seedrate-41.92958)/10
site a numeric factor indexing the different location by year combinations
yield yields (tonnes/hectare)
There are 1420 lines of data.
ELISA data: calibration of the protein DNase. The data file contains 4 variables on
each line, with the top row labelling the variables, namely
y the optical density measurements
fc a factor for the concentration levels
conc concentration values
Trial the trial as a factor with 11 levels
There are 176 lines of data.
|
