Defines options for the fitting of models by
VARANDOM and associated procedures (R.W. Payne).
||Limit on the number of iterations in
||Number of blocks of internal memory to be set up for use by the
||Minimum number of different coordinates in a direction for a spatial model to be fitted by
||Critical value for the probability of a row or column trend in the initial basic
||Whether the accumulated summary should include models that fail to fit or that have bound variance parameters (
There are several procedures, with the prefix
VA, that investigate potential models for a
REML analysis. For example,
VABLOCKDESIGN does this for data from an incomplete-block design, and
VAROWCOLUMNDESIGN does it for data from a field trial arranged in rows and columns. These two procedures both use a general procedure
VARANDOM, which allows you to try several alternative random models for a
REML analysis, and then select the best one according to either their Akaike or Schwarz (Bayesian) information coefficients.
VAROWCOLUMNDESIGN are designed to be used by non-expert users, and so various decisions are made there (and in
VARANDOM) about how the analysis is to be done and how the models are to be selected. This procedure,
VAOPTIONS, is provided to allow more experienced users to modify some of the options that control the process.
WORKSPACE options are relevant to all the procedures, and define the settings for the
WORKSPACE options of the
REML directive, when this is used in any of the procedures.
Other options are relevant to specific procedures. The
LIMPRTREND options control aspects of the analyses in
MINSPATIALCOORDINATES sets a limit on the number of different row or column coordinates for a spatial model to be fitted in that direction (default 4).
LIMPRTREND defines the critical value for the probability of a row or column trend in an initial basic
REML analysis, that is performed with replicates but no other random terms, to decide whether this is be included in the later analyses (default 0.01). The
REPORTFAILURES option controls whether the accumulated summary, printed from
VARANDOM, should include models that fail to fit or that have bound variance parameters.
Commands for: REML analysis of linear mixed models.
CAPTION 'VAOPTIONS example',\ 'Slate Hall Farm data (Guide to REML in Genstat, Section 1.8).';\ STYLE=meta,plain SPLOAD '%gendir%/data/slatehall.gsh' " define model for analysis as a randomized-blocks design " VFMODEL [MODELSTRUCTURE=RCBD; DESCRIPTION='Randomized blocks';\ FIXED=variety] replicates " define model for analysis as a Lattice square design " VFMODEL [MODELSTRUCTURE=Latticesq; DESCRIPTION='Lattice square';\ FIXED=variety] replicates/(rows*columns) " define model for analysis with an Ar1 (x) Ar1 model " VFMODEL [MODELSTRUCTURE=AR1xAR1; DESCRIPTION='Ar1 (x) AR1';\ FIXED=variety] fieldrow.fieldcolumn VFSTRUCTURE [MODELSTRUCTURE=AR1xAR1; TERMS=fieldrow.fieldcolumn]\ 2('AR'); ORDER=1; FACTOR=fieldrow,fieldcolumn " define model for analysis with an Ar1 (x) Ar1 model + measurement error " VFMODEL [MODELSTRUCTURE=AR1xAR1p; DESCRIPTION='Ar1 (x) AR1 + plots';\ FIXED=variety] fieldrow.fieldcolumn+plotnumber VFSTRUCTURE [MODELSTRUCTURE=AR1xAR1p; TERMS=fieldrow.fieldcolumn]\ 2('AR'); ORDER=1; FACTOR=fieldrow,fieldcolumn " find the best random model " VARANDOM [PTRY=monitor; MODELSTRUCTURES=RCBD,Latticesq,AR1xAR1,AR1xAR1p]\ yield; SAVE=savebest " set MAXCYCLE=8 in future REMLs (the spatial analyses will now fail) " VAOPTIONS [MAXCYCLE=8] VARANDOM [PTRY=monitor; MODELSTRUCTURES=RCBD,Latticesq,AR1xAR1,AR1xAR1p]\ yield; SAVE=savebest " restore MAXCYCLE to the default 100 " VAOPTIONS [MAXCYCLE=100]