Constructs an augmented block design, using CycDesigN if the controls are in an incomplete-block design (R.W. Payne).
Options
PRINT = strings |
Controls printed output (design, controldesign, factors, monitor); default * i.e. none |
|---|---|
LEVELS = scalar or variate |
Levels for the unreplicated treatments |
LEVCONTROLS = scalar or variate |
Levels for the control treatments |
NROWS = scalar |
Number of rows |
NCOLUMNS = scalar |
Number of columns |
NRBLOCKS = scalar |
Number of rows in each block |
NCBLOCKS = scalar |
Number of columns in each block |
NCONTROLSPERBLOCK = scalar |
Number of control treatments in each block |
TREATMENTS = factor |
Treatment factor |
ROWS = factor |
Row factor |
| COLUMNS = factor | Column factor |
BLOCKS = factor |
Block factor |
ROWBLOCKS = factor |
Row block factor |
COLBLOCKS = factor |
Column block factor |
NTIMES = scalar |
Number of times to try allocations of controls within blocks |
SEED = scalar or variate |
Scalar or variate with three values specifying seeds for the random numbers used by CycDesigN to search for the control design, for the allocation of controls within blocks, and for the allocation of the unreplicated treatments – if a scalar is specified the same seed is used for all purposes; default 0 i.e. set automatically |
SPREADSHEET = string |
Whether to put the design factors into a spreadsheet (design); default * |
TIMELIIMIT = scalar |
Time in minutes for CycDesigN to search; default 1 |
No parameters
Description
An augmented design starts with a basic design containing control treatments. This is then augmented by adding extra plots for the unreplicated test treatments. The basic design in CDNAUGMENTEDDESIGN is a block design, where the blocks can be in either a one- or a two-dimensional layout. The augmenting stage expands the blocks with extra plots to contain the unreplicated treatments.
The levels of the treatment factor to be used for the unreplicated treatments are specified by the LEVELS option. This can be a scalar, to specify levels 1, 2 etc., or a variate specifying the actual levels to use. The levels for the control treatments are specified similarly by the LEVCONTROLS option. There must be more than two controls, and their levels must be distinct from those for the unreplicated treatments.
The NROWS option can define the number of rows in the design, and the ROWS option can supply a factor to save the levels generated for the row factor. You can omit NROWS if ROWS is set to a factor that has already been defined with the correct number of levels. Similarly, the NCOLUMNS option can define the number of columns, and the COLUMNS option can supply a factor to save the levels generated for the column factor.
The NRBLOCKS and NCBLOCKS options specifies the number of rows and columns, respectively, to be used for each block. The numbers of rows and columns in the design must be exact multiples of NRBLOCKS and NCBLOCKS. You can specify a one-dimensional layout for the blocks by setting either NRBLOCKS equal to the number of rows in the design, or NCBLOCKS equal to the number of columns in the design. Otherwise the blocks are in a two-dimensional (row-by-column) layout. The BLOCKS option can save a factor containing the levels generated for the blocks. Also, the ROWBLOCKS option can save a factor containing the row location of each block in the two-dimensional (row-by-column) layout, the and the COLBLOCKS option can save a factor containing its column location.
The NCONTROLSPERBLOCK option specifies the number of plots for controls in each block. If this is less than the number of controls, CycDesigN is used to find an efficient incomplete-block design for the controls. (Otherwise the basic design is a randomized complete block design and CycDesigN is not needed.) The TIMELIMIT option specifies the time in minutes for CycDesigN to search.
CDNAUGMENTEDDESIGN then tries several randomizations for the controls within the blocks, and takes the one that has the most uniform allocation of the control treatments over the rows and columns within the blocks. The NTIMES option specifies the number of randomizations to try (default 1000).
The SEED option allows you to supply seeds for the random numbers to be used for the random numbers used by CycDesigN to search for the control design, for the allocation of controls within blocks, and for the allocation of the unreplicated treatments . You can specify a variate with three values to supply a different seed for each purpose, or a scalar to use the same one for both. If a zero value is specified, the corresponding seed is set automatically. The default is the scalar zero.
Printed output is controlled by the PRINT option, with settings:
design to print the design,
controldesign to print the design showing just the control treatments,
factors to print the factor values, and
monitor to print a report by CycDesigN on the design and monitor the randomizations of the control design.
By default nothing is printed.
You can set option SPREADSHEET=design to put the design factors into a Genstat spreadsheet.
Options: PRINT, LEVELS, LEVCONTROLS, NROWS, NCOLUMNS, NRBLOCKS, NCBLOCKS, NCONTROLSPERBLOCK, TREATMENTS, ROWS, COLUMNS, BLOCKS, ROWBLOCKS, COLBLOCKS, NTIMES, SEED, SPREADSHEET, TIMELIMIT.
Parameters: none.
Method
The batch CycDesigN program is called using the SUSPEND directive.
See also
Procedures: AFAUGMENTED, CDNPREP, CDNBLOCKDESIGN, CDNROWCOLUMNDESIGN.
Commands for: Design of experiments.
Example
CAPTION 'CDNAUGMENTEDDESIGN example',\
!t('Augmented design for 120 genotypes and 4 controls',\
'in a 3 row x 2 column array of blocks'); STYLE=meta,plain
CDNAUGMENTEDDESIGN [PRINT=design; LEVELS=!(101...220); LEVCONTROL=4; \
NROWS=12; NCOLUMNS=12; NRBLOCKS=4; NCBLOCKS=6; TREATMENT=genotypes; \
ROWS=rows; COLUMNS=columns; BLOCKS=blocks; SEED=1274410]