Blog Archives

New in forecast 5.0

January 26, 2014
By
New in forecast 5.0

Last week, version 5.0 of the forecast package for R was released. There are a few new functions and changes made to the package, which is why I increased the version number to 5.0. Thanks to Earo Wang for helping with this new version. Handling missing values and outliers Data cleaning is often the first step that data scientists...

Read more »

Thoughts on the Ljung-Box test

January 23, 2014
By
Thoughts on the Ljung-Box test

It is common to use a Ljung-Box test to check that the residuals from a time series model resemble white noise. However, there is very little practical advice around about how to choose the number of lags for the test. The Ljung-Box test was proposed by Ljung and Box (Biometrika, 1978) and is based on the statistic    ...

Read more »

Slides from my online forecasting course

January 22, 2014
By

Last year I taught an online course on forecasting using R. The slides and exercise sheets are now available at www.otexts.org/fpp/resources/

Read more »

Looking for a new post-doc

January 21, 2014
By

We are looking for a new post-doctoral research fellow to work on the project “Macroeconomic Forecasting in a Big Data World”.  Details are given at the link below jobs.monash.edu.au/jobDetails.asp?sJobIDs=519824 This is a two year position, funded by the Australian Research Council, and working with me, George Athanasopoulos, Farshid Vahid and Anastasios Panagiotelis. We are looking for someone with a PhD...

Read more »

Estimating a nonlinear time series model in R

January 20, 2014
By
Estimating a nonlinear time series model in R

There are quite a few R packages available for nonlinear time series analysis, but sometimes you need to code your own models. Here is a simple example to show how it can be done. The model is a first order threshold autoregression:     where is a Gaussian white noise series with variance . The following function will generate...

Read more »

Judgmental forecasting experiment

December 22, 2013
By

The Centre for Forecasting at Lancaster University is conducting some research on judgmental forecasting and model selection. They hope to compare the performance of judgmental model selection with statistical model selection, in order to learn how to best design forecasting support systems. They would like forecasting students, practitioners and researchers to participate, and are offering £50 Amazon Gift Cards...

Read more »

Three jobs at Monash

October 17, 2013
By

We are currently advertising for three academic positions, suitable for recent PhD graduates. Lecturer (Applied Statistics or Operations Research) Five-year position with MAXIMA and the School of Mathematical Sciences Two positions available. Applications close 31 October. More information. Lecturer (Econometrics/Business Statistics) Continuing position with the Department of Econometrics and Business Statistics Applications close 31 January 2014. More information. Please...

Read more »

Questions on my online forecasting course

October 3, 2013
By

I’ve been getting emails asking questions about my upcoming course on Forecasting using R. Here are some answers. Do I need to use the Revolution Enterprise version of R, or can I use open-source R? Open source R is fine. Revolution Analytics is organizing the course, but there is no requirement to use their software. I will be using...

Read more »

Forecasting with R

September 25, 2013
By

The following video has been produced to advertise my upcoming course on Forecasting with R, run in partnership with Revolution Analytics. The course will run from 21 October to 4 December, for two hours each week. More details are available at http:/...

Read more »

Forecasting with daily data

September 16, 2013
By
Forecasting with daily data

I’ve had several emails recently asking how to forecast daily data in R. Unless the time series is very long, the simplest approach is to simply set the frequency attribute to 7. y <- ts(x, frequency=7) Then any of the usual time series forecasting methods should produce reasonable forecasts. For example library(forecast) fit <- ets(y) fc <- forecast(fit) plot(fc)...

Read more »