The VARIOGRAM Procedure

Syntax: VARIOGRAM Procedure

The following statements are available in PROC VARIOGRAM:

PROC VARIOGRAM options ;: BY variables ;; COMPUTE computation-options ;; COORDINATES coordinate-variables ;; DIRECTIONS directions-list ;; ID variable ;; MODEL model-options ;; PARMS parameters-list < / parameters-options> ;; NLOPTIONS <options> ;; STORE store-options ;; VAR analysis-variables-list ;

The COMPUTE and COORDINATES statements are required. The MODEL and PARMS statements are hierarchical. If you specify a PARMS statement, it must follow a MODEL statement.

Table 98.1 outlines the options available in PROC VARIOGRAM classified by function.

Table 98.1 Options Available in the VARIOGRAM Procedure
Task	Statement	Option
Data Set Options
Specify input data set	PROC VARIOGRAM	DATA=
Suppress normal display of results	PROC VARIOGRAM	NOPRINT
Write autocorrelation weights information	PROC VARIOGRAM	OUTACWEIGHTS=
Write distance histogram information	PROC VARIOGRAM	OUTDISTANCE=
Write Moran scatter plot information	PROC VARIOGRAM	OUTMORAN=
Write pairwise point information	PROC VARIOGRAM	OUTPAIR=
Write spatial continuity measures	PROC VARIOGRAM	OUTVAR=
Specify the plot display and options	PROC VARIOGRAM	PLOTS
Specify a model data set with MODEL statement	MODEL	MDATA=
Specify a model data set with PARMS statement	PARMS	PDATA=
Declaring the Role of Variables
Specify variables to define analysis subgroups	BY
Specify variable with observation labels	ID
Specify the analysis variables	VAR
Specify the x, y coordinates in the DATA= data set	COORDINATES	XCOORD= YCOORD=
Controlling Continuity Measure Computations
Specify the confidence level	COMPUTE	ALPHA=
Specify the angle tolerances for angle classes	COMPUTE	ANGLETOLERANCE=
Compute autocorrelation statistics	COMPUTE	AUTOCORRELATION
Specify the bandwidths for angle classes	COMPUTE	BANDWIDTH=
Compute the semivariance estimate variance	COMPUTE	CL
Specify the minimum distance that indicates any two distinct points are not collocated	COMPUTE	DEPSILON=
Specify the basic lag distance	COMPUTE	LAGDISTANCE=
Specify the tolerance around the lag distance	COMPUTE	LAGTOLERANCE=
Specify the maximum number of lags in computations	COMPUTE	MAXLAGS=
Specify the number of angle classes	COMPUTE	NDIRECTIONS=
Suppress computation of all continuity measures	COMPUTE	NOVARIOGRAM
Compute robust semivariance	COMPUTE	ROBUST
Controlling Distance Histogram Data Set
Specify the distance histogram data set	PROC VARIOGRAM	OUTDISTANCE=
Specify the number of histogram classes	COMPUTE	NHCLASSES=
Controlling Pairwise Information Data Set
Specify the pairwise data set	PROC VARIOGRAM	OUTPAIR=
Specify the maximum distance for the pairwise data set	COMPUTE	OUTPDISTANCE=
Controlling Semivariogram Model Fitting
Specify the item store to save correlation information	STORE	OUT=
Specify the confidence level for fitting parameters	MODEL	ALPHA=
Specify fitted model ranking criteria	MODEL	CHOOSE=
Compute parameters estimate limits	MODEL	CL
Specify a threshold to compare model fit quality	MODEL	RANKEPS=
Specify a tolerance to use in model classification	MODEL	EQUIVTOL=
Specify the type of semivariogram to fit	MODEL	FIT=
Specify a type with a functional form	MODEL	FORM=
Specify the model fitting method	MODEL	METHOD=
Specify a minimal nugget effect if experimental semivariance is zero at first lag	MODEL	NEPSILON=
Suppress model fitting	MODEL	NOFIT
Specify the nugget effect for fitted model	MODEL	NUGGET=
Specify a range estimate for fitted model	MODEL	RANGE=
Specify a range of lags to fit a model in	MODEL	RANGELAG=
Specify a scale estimate for fitted model	MODEL	SCALE=
Specify a Matérn smoothness estimate	MODEL	SMOOTH=
Specify constant parameters in fitting	PARMS	HOLD=
Specify fitting parameter lower bounds	PARMS	LOWERB=
Specify the upper limit for fitted scale	PARMS	MAXSCALE=
Specify no bounds for fitted parameters	PARMS	NOBOUND
Specify the fitting parameter upper bounds	PARMS	UPPERB=
Specify optimization process options	NLOPTIONS
Fitting Output Tables Control Options
Request the approximate covariance matrix	MODEL	COVB
Request the approximate correlation matrix	MODEL	CORRB
Request fit details for every candidate model	MODEL	DETAILS
Request the gradient of the objective function in parameter estimates table	MODEL	GRADIENT
Threshold to switch a Matérn form to Gaussian	MODEL	MTOGTOL=
Suppress the iteration history table	MODEL	NOITPRINT