Command File
Context: Syntax
Purpose
ASReml is essentially a batch program controlled with a command file.
The command file normally has a filename extension
.as.
Which editor
You will need
an ASCII file editor to create the command file and to view the various output files.
On unix systems,
vi
and
emacs
are commonly used.
ASReml-W
is also available.
Under Windows,
there are several suitable program editors available such as
ASReml-W
and
ConText.
General rules
In the ASReml command file
all characters following a
#
symbol on a line are ignored,
all blank lines are ignored,
lines beginning with
!
followed by a blank are copied to the
.asr
file as comments for the output,
a blank is the usual separator; TAB is also a separator,
maximum line length is 2000 characters,
the
//
operator splits a line into
two logical lines,
a comma as the last character on the line is used to
indicate that the current list is continued on the next line; a
comma is not needed when ASReml knows how many values to read,
reserved words used in specifying the linear model
are case sensitive; they need to be
typed exactly as defined: they may not be abbreviated.
a qualifier is a particular letter sequence beginning with an
!
which sets an option or changes some aspect of ASReml;
some qualifiers require arguments,
qualifiers must appear on the correct line,
qualifier identifiers are not case sensitive,
qualifier identifiers may be truncated to 3 characters.
A minimal job
Every commandfile has at least four lines, in the order
Title line - gives a heading for the job
data field names - as they occur in the data file (indent these lines)
name of the data file
model line
Example
NIN alliance trial 1989 # Title line
variety !A # names of data fields
id
pid
raw
repl 4
nloc
yield
lat
long
row 22
column 11
nin89.asd !skip 1 !DOPATH $1 # datafile line
!PATH 1 # RCB analysis
tabulate yield ~ variety # request simple variety means
yield ~ mu variety !r repl # Model Line
predict variety # request predicted variety means
!PATH 2 # Spatial Analysis
yield ~ mu variety
predict variety
1 1 # Variance header
row row AR1 0.5
column column AR1 0.5
Return to start