Loading ...
Sorry, an error occurred while loading the content.


Expand Messages
  • Keys Deborah DA (AG Prod, WRC)
    Mar 29, 1999

      Is anyone familiar with SAS PROC MIXED??

      I am using it to fit semivariograms to spatially correlated data and
      designed experiment data with spatially correlated errors.

      If your answer is yes.
      Here is my technical question.

      In SAS v 6.12 (current version), when the sill, range and nugget are
      estimated as a spatial covariance structure (by using the command LOCAL in
      the REPEATED statement) which covariance parameter estimates corresponds to
      which parameter (range, sill or nugget) on the SAS output,

      Here are 2 examples for spherical and gaussian covariance structures, For
      the first one it makes sense that the estimates are nugget, range, sill (top
      to bottom). But if this is true, in the second example the nugget would be
      larger than the sill but this can't happen!!

      Example 1
      Covariance Parameter Estimates (REML)

      Cov Parm Subject Estimate

      Variance INTERCEPT 0.43177363
      SP(SPH) INTERCEPT 54.78167843
      Residual 1.04333638

      Example 2
      Covariance Parameter Estimates (REML)

      Cov Par Subject Estimate
      Variance INTERCEPT 2.07463337
      SP(GAU) INTERCEPT 252.65941634
      Residual 1.25734694

      And a follow up question, Does SAS estimate the sill as an absolute value on
      the y axis or as an increment above the nugget and the value of the max

      Thanks. I will post any answers I receive.

      Deborah Keys

      Quantitative Agronomist
      Zeneca Ag Products
      *(510) 231-1514

      *To post a message to the list, send it to ai-geostats@....
      *As a general service to list users, please remember to post a summary
      of any useful responses to your questions.
      *To unsubscribe, send email to majordomo@... with no subject and
      "unsubscribe ai-geostats" in the message body.
      DO NOT SEND Subscribe/Unsubscribe requests to the list!
    • Show all 2 messages in this topic