Basic Information

  • Instructor: Mikyoung Jun (webpage, mjun {at} stat {dot} tamu {dot} edu)
  • Classes meet: TR 12:45-2:00 PM at room BLOC 411
  • Office hours: TR 2:00-3:00 PM or by appointment
  • Text books: there is no specific text book but here is a list of books that course materials will be based on:
  • Statistical Methods for Spatial data Analysis by Schabenberger and Gotway, 2005; CRC Press
  • Statistics for spatial data by Cressie, N, 1993; Wiley
  • Course description: this course will cover the following topics (not necessarily the same order nor equal emphasis)
  • Auto correlation
  • Stationary, isotropic random fields
  • Point processes
  • Variograms
  • Kriging
  • Estimation methods for covariance parameters
  • Change of support problem
  • Spatial regression
  • Simulation methods
  • Nonstationary processes
  • Spatial-temporal processes
  • Processes on a sphere
  • Some Bayesian methods in spatial statistics
  • Course grade: homework assignments (20%), one midterm (30%), team project (50%) (tentative plan)
  • Course syllabus: here
  • Course schedule: here is a list of tentative course schedule.

    • Important Announcements

  • There will be no class on November 24. Spend this time for the final project.
  • Midterm will be on October 29, Thursday, NOT on October 27, Tuesday. October 27, Tuesday will be a reading day since I will be occupied by a workshop on that day.
  • There will be no class on September 15th (I will be away for a conference)

    • Course Material

  • data 1
  • lecture 1
  • lecture 2
  • lecture 3
  • lecture 4
  • lecture 5
  • data 2
  • lecture 6
  • lecture 7 (corrected)
  • lecture 8
  • data 3
  • lecture 9
  • lecture 10
  • lecture 11
  • lecture 12
  • lecture 13
  • lecture 14
  • lecture 15
  • data 4
  • lecture 16
  • lecture 17
  • lecture 18
  • lecture 19
  • lecture 20
  • lecture 21

    • Homework Solutions

  • HW1 (by Ganggang Xu)
  • HW2 (by Xinxin Zhu)
  • HW3 (part 1 by Xinxin Zhu)
    • : many of you made several mistakes in displaying variograms. Be careful in calculating the distances. Here the distance should be naturally arc distance since data are on a sphere and the coordinate is lon/lat. Also the nonstationarities that you may see in the variograms should be mainly due to the mean in the data. Once you filter the mean out, you shouldn't see much nonstationarity in the data. Actually when I simulated the data set, I used an isotropic covariance model.
  • HW3 (part 2 by Ho Jung Yoon)
  • HW4 (written by me)
  • HW 4 (witten by Sean Tolle) homework  R code
  • HW5 (written by me)
  • HW6 (written by me)


    • Recorded lectures (only for registerd students)

    link

    Last modified: 11/19/09