Forms a variate and set of classifying factors from a table (P.W. Goedhart).
No options
Parameters
TABLE = tables |
Tables to be copied |
|---|---|
VARIATE = variates |
New variate to contain the body of each table |
CLASSIFICATION = pointers |
Pointer containing the factors by which each new variate is classified |
LABELS = texts |
Labels for the new variate, indicating the values of the classifying factors corresponding to each of its units |
Description
This procedure can be used to store the body of a table in a variate and obtain a set of factors to represent the way in which the data are arranged in the table. These factors then classify the newly formed variate in the same way as in the table. You can also form a text containing labels formed from the values of the classifying factors. Margins of the table are ignored.
The table to be copied is specified by the TABLE parameter, the variate must be specified by the VARIATE parameter, the set of classifying factors can be obtained by setting CLASSIFICATION to a pointer, and the labels by setting the LABELS parameter to a text. If the CLASSIFICATION pointer has not been declared, the names of the classifying factors of the table are used as suffix names. The newly formed factors have the same attributes as the old classifying factors, excluding the setting of EXTRA. Note that the order in which the factors are obtained can be unexpected for implicitly declared tables as explained in the Guide to Genstat, Part 1, Section 4.1.5.
Options: none.
Parameters: TABLE, VARIATE, CLASSIFICATION, LABELS.
Method
Margins of the table are deleted by the directive MARGIN. The classifying factors of the table are obtained with GETATTRIBUTE. The initial declarations of the new factors are done by DUPLICATE to transfer any relevant attributes. Factor values are then produced by GENERATE.
See also
Procedure: FBETWEENGROUPVECTORS, VMATRIX.
Commands for: Calculations and manipulation.
Example
CAPTION 'VTABLE example',!t(\
'In a randomized block experiment with 4 blocks of 3 units each,',\
'the effect of 3 different diets on milk production was tested.',\
'Each unit consisted of 4 or 5 cows; milk production was measured',\
'individually. VTABLE is used to preprocess the data for an',\
'analysis of variance.'); STYLE=meta,plain
FACTOR [LEVELS=4 ; VALUES=12(1),15(2),13(3),14(4)] Block
FACTOR [LABELS=!T('Nitrogen+','Nitrogen0','Nitrogen-') ;\
VALUES=4(1,2,3), 5(1,2,3), 4(1,2,3),3, 5(1,2),4(3)] Diet
VARIATE [NVALUES=54] Milk
READ Milk; DECIMALS=1
312 330 300 287 294 291 303 289 275 282 281 290
278 284 281 263 289 294 283 281 274 298 264 270 288 285 248
290 256 265 243 270 261 256 279 253 259 268 240 242
276 243 233 238 259 245 241 227 255 222 235 227 227 247 :
TABULATE [CLASSIFICATION=Block,Diet] Milk ; MEAN=TabMilk
PRINT TabMilk; DECIMALS=1
VTABLE TABLE=TabMilk; VARIATE=MeanMilk;\
CLASSIFICATION=!P(MeanBlock, MeanDiet)
PRINT MeanMilk,MeanBlock,MeanDiet; DECIMALS=1,2(*)
BLOCK MeanBlock
TREAT MeanDiet
ANOVA MeanMilk