Department of Statistics

Stanford University 

Summer 2001


Stat 207: Introduction to Time Series Analysis

Course Description Homework Important Dates Statistics Software
Syllabus Handouts Books Statistics Online Texts/Courses


Week 8:

  • The final is on Friday August 17, 12:15 - 3:15.
  • Homework #6 is due of Friday.
  • The Project is due on Friday.
  • This week we will finish some remaining topics related to Frequency Domain Analysis, we will discuss signal detection, and have a brief introduction to State Space Models.

Week 7:

  • This week we will cover Frequency Domain analysis.
  • Homework #6 has been assigned.
  • Two handouts related to Chapter 3 have been posted.

Week 6:

  • The solution to the in-class midterm is now available.
  • This week we will finish Estimation, we will discuss Regression with Auto-correlated Errors and Transfer Function Modeling, and begin the Frequency Domain Analysis of Time Series Data.
  • Homework #5 has been assigned.
  • You should begin looking for a dataset of interest to you to analysis for the class project.  You might consider looking at these two websites.  
  • For problem 2.36 on the homework, if you are using R you will need to download and install the fracdiff library.  To install a library you need to download the file containing the library from the R-project contributed packages website and then run the install program selecting where the library file is and then where you want to install it.  The directory you should install to in the \R\library\ directory.  To call the library, library(fracdiff).  I have posted an example on the Handouts page.
  • For problems 2.39 which is about regression with auto-correlated errors, I suggest that you use ASTSA's Multiple Regression option under Time Domain and the ARIMA residual option under Transform and then Transform.  This problem can also be done using Splus, using the arima.filt( ) function, or using R, using the filter( ) funtion.
  • For problems 2.41 and 2.42 which are related to transfer function modeling, I suggest that you use ASTSA and the options mentioned above.  The data file is clim-hyd.asd.

Week 5:

  • The In-class midterm is on Tuesday.  I plan to have it graded so it can be returned on Thursday.
  • The Take-home Midterm has three datasets: racife.dat securities.dat airpass.dat
  • There are also two handouts related to performing the Classical Decomposition of a Racife, Brazil Air Temperature data. racife.r racife.ssc
  • The fourth homework assignment and the Take-home Midterm are both due on Monday next week.

Week 4:

  • This week we will cover model ARIMA development and selection, forecasting, and estimation.
  • See Homework for the simulated datasets for Problems 2.6 and 2.8.
  • See Handouts for more Splus examples.
  • See Homework for the solutions to Homework 1 and 2.
  • The midterm will be next Tuesday in class and there will be a take-home part of the midterm that will be due the following Monday.

Week 3:

  • This week we will introduce the general ARMA model.
  • See Handouts for an Splus code that simulated ARIMA data and plots the ACF's and PACF's and fits ARIMA models to the simulated data.

Week 2:

  • Introduction to time domain modeling.
  • The datasets from the book are available from both of the author's websites in dos format and in unix format.  See the Books above link also.
  • See Handouts for more Splus examples.

Week 1:

  • I plan to be on campus Monday June 25 in the afternoon if anyone has any questions.
  • Class begins on June 26.

Week 0:

  • The book I have chosen for the course is Shumway and Stoffer: Time Series Analysis and Its Applications, Springer Verlag, 2000.