Category Archives: Time series

What’s wrong with LOESS for palaeo data?

Locally weighted scatterplot smoothing (LOWESS) or local regression (LOESS) is widely used to highlight “signal” in variables from stratigraphic sequences. It is a user-friendly way of fitting a local model that derives its form from the data themselves rather than having … Continue reading

Posted in Palaeoecology, Palaeolimnology, R, Science, Time series | Tagged , , , , | 4 Comments

Smoothing temporally correlated data

Something I have been doing a lot of work with recently are time series data, to which I have been fitting additive models to describe trends and other features of the data. When modelling temporally dependent data, we often need … Continue reading

Posted in R, Time series | 4 Comments

Embedding a time series with time delay in R — Part II

Some months ago, I posted a function that extended the base R function embed() to allow for time delay embedding. Today, David Gonzales alerted me to an inconsistency between embed() and Embed(). The example David used was where Embed() clearly … Continue reading

Posted in R, Time series | Tagged , | 2 Comments

Additive modelling and the HadCRUT3v global mean temperature series

Earlier, I looked at the HadCRUT3vgl data set using generalized least squares to investigate whether the trend in temperature since 1995 was statistically significant. Here I want to follow-up one of the points from the earlier posting; namely using a … Continue reading

Posted in Climate Change, R, Science, Time series | 16 Comments

Global warming since 1995 ‘now significant’

Updated: 17 July 2011 Yesterday (June 11, 2011) the BBC reported on comments by Prof. Phil Jones, of the Climatic Research Unit (CRU), University of East Anglia (UEA), that the warming experienced by the planet since 1995 was statistically significant. … Continue reading

Posted in Climate Change, R, Science, Time series | 17 Comments

Embedding a time series with time delay in R

I’ve recently been looking at Martin Trauth‘s book MATLAB® Recipes for Earth Sciences to try to understand what some of my palaeoceanography colleagues are doing with their data analyses (lots of frequency domain time series techniques and a preponderance of … Continue reading

Posted in R, Time series | 3 Comments