FOCUS AREAS

SAS/STAT Topics

SAS/STAT Software

Spatial Analysis

The goal of spatial data analysis is to derive insights from the location and context of real world phenomena such as crimes, accidents, vegetation growth, availability of minerals, and so on. Spatial analysts are often concerned with why things happen where they do. They are also concerned with how the occurrence of a phenomenon in a particular region affects nearby regions. More precisely, the “where” in spatial data analysis usually refers to one of three possibilities:

Application areas of spatial analysis include but not limited to, agriculture, forestry, crime analysis, public health, environmental monitoring, insurance, and, energy.

The SAS/STAT spatial analysis procedures include the following:

KRIGE2D Procedure


The KRIGE2D procedure performs ordinary kriging or spatial prediction for spatial point referenced data. The following are highlight's of the procedure's features:

  • enables you to specify the correlation model by naming the form and supplying the associated parameters, or by using the contents of an item store file that was previously created by PROC VARIOGRAM
  • enables you to specify the locations of kriging predictions in a GRID statement, or they can be read from a SAS data set
  • local kriging is supported through the specification of a radius around a grid point or the specification of the number of nearest neighbors to use in the kriging system
  • produces an output data set that contains the kriging predictions and associated standard errors for each grid
  • produces an output data set that contains the neighborhood information for each grid point when you perform local kriging
  • supports BY group processing, which enables you to obtain separate analyses on grouped observations
  • automatically creates graphs by using ODS Graphics
For further details, see KRIGE2D Procedure

SIM2D Procedure


The SIM2D procedure produces a spatial simulation for a Gaussian random field with a specified mean and covariance structure in two dimensions by using an LU decomposition technique. The simulation can be conditional or unconditional. The following are highlights of the SIM2D procedure's features:

  • enables you to specify the mean structure as a quadratic function in the coordinates
  • enables you to specify the covariance by naming the form and supplying the associated parameters, or by using the contents of an item store file that was previously created by PROC VARIOGRAM
  • enables you to specify the locations of simulation points in a GRID statement, or they can be read from a SAS data set
  • performs BY group processing, which enables you to obtain separate analyses on grouped observations
  • writes the simulated values for each grid point to an output data set
  • creates a SAS data set that corresponds to any output table
  • automatically creates graphs by using ODS Graphics
For further details, see SIM2D Procedure

SPP Procedure


The SPP procedure analyzes spatial point patterns. The broad goal of spatial point pattern analysis is to describe the occurrence of events (observations) that compose the pattern. The event locations are a discrete realization of a random spatial process. Therefore, the analysis goal is to investigate and characterize the original spatial process that generated the events in the spatial point pattern. The procedure's capabilities include the following:

  • fits an inhomogeneous Poisson process model and produce a variety of residual diagnostics
  • performs a simulation based goodness-of-fit test for the fitted intensity model
  • produces a nonparametric estimate of the first-order intensity
  • computes the F, G, J, K, L, and pair correlation functions and performs a test for complete spatial randomness based on these functions
  • performs a quadrat based test for complete spatial randomness
  • performs marked point pattern analysis by using character mark variables
  • automatically computes the study area bounds for the spatial point pattern by using the Ripley-Rasson estimator
  • performs covariate dependency tests that are based on an empirical distribution function
  • manages duplicate records
  • performs BY group processing, which enables you to obtain separate analyses on grouped observations
  • creates a SAS data set that corresponds to any output table
  • automatically creates graphs by using ODS Graphics
For further details, see SPP Procedure

VARIOGRAM Procedure


The VARIOGRAM procedure computes variogram diagnostics to describe the spatial covariance structure in spatial point referenced data. The following are highlights of the VARIOGRAM procedure's features:

  • handles anisotropic and nested semivariogram models
  • the following eight semivariogram models are supported:
    • Gaussian
    • exponential
    • spherical
    • power
    • cubic
    • pentaspherical
    • sine hole effect
    • Matern
  • supports a single nugget effect
  • fits permissible theoretical models to the empirical semivariograms, so that you can use them in subsequent analysis to perform spatial prediction
  • produces plots of the empirical semivariograms and the fitted models
  • provides isotropic and anisotropic measures
  • provides the Moran I and Geary c spatial autocorrelation statistics
  • enables you to save the context and results of the semivariogram model fitting analysis in an item store
  • performs BY group processing, which enables you to obtain separate analyses on grouped observations
  • creates a SAS data set that corresponds to any output table
  • automatically creates graphs by using ODS Graphics
  • For further details, see VARIOGRAM Procedure