Richard Telford’s Blog
@richardjtelfordRecent Comments

Recent Posts
Archives
Categories
Tags
analogue quality autocorrelation Baffin Island Barycentre Benthic foraminifera Bob Irvine Bulafu et al (2013) Chen et al (2013) Chen et al. (2011) chironomids climate models colours critical transition data archiving diatoms Dietl (2016) dinocysts Di Rita (2013) Doug Keenan EDA Engels and van Geel (2012) graphics Guiot and de Vernal (2011) hblock crossvalidation Heartland impact INTIMATE training school Kelsey et al 2015 Klein et al (2013) Klemm et al 2013 LarocqueTobler et al (2015) LGM Liu et al 2014 Lyons et al 2016 maps Marcott et al 2013 Mauri et al (2015) Miller et al 2013 moss Murry Salby NIPCC NMDS Norway ordination palaeoSig Peer review Planktonic foraminifera Poland pollen R radiocarbon calibration radiocarbon dating Reconstruction diagnostics reconstructions REDFIT Rhine freezing Salicornia Schmittner et al. (2011) Schulz & Mudelsee (2002) Sea ice sea ice reconstruction Sirocko et al (2012) Skjærvø et al (2015) social network analysis spatial autocorrelation Spectral analysis Tatra Mountains Telford (2006) Telford and Birks (2011) Transfer function diagnostics Uganda uneven sampling van Oldenborgh et al. (2013) Willie Soon Younger DryasBlogroll
Follow me on Twitter
My TweetsMeta
 Follow Musings on Quantitative Palaeoecology on WordPress.com
Category Archives: EDA
Falling for ggplot2
I spent a long time resisting the lure of ggplot2. I was proficient with the plotting functions in base graphics; why did I need to learn an entirely new graphics system? Yes, getting up colour ramps could be a real … Continue reading
Posted in EDA, R
Leave a comment
Variance inflation factors and ordination model selection
Variance inflation factors (VIF) give a measure of the extent of multicollinearity in the predictors of a regression. If the VIF of a predictor is high, it indicates that that predictor is highly correlated with other predictors, it contains little or no unique information, and there is redundancy … Continue reading
REDFIT’s rule of thumb
Because REDFIT tests many frequencies, some are likely to appear statistically significant just by chance — a classic multiple testing problem. Schulz & Mudelsee (2002) “follow Thomson (1990) and select a falsealarm level of (11/n)*100%, where n is the number of data points … Continue reading
REDFIT & false alarms
REDFIT is a useful tool for palaeoecologists who like to test their data for periodicities as it uses the LombScargle Fourier transform which tolerates unequal time intervals and so avoids the problems inherent in interpolating data to equal intervals. Several of the papers reporting the … Continue reading
Collaboration networks in BIO
I’ve been intrigued about social network analysis since reading the Wegman report that seemed to find that Michael Mann was at the centre of the network of coauthors on papers on which he was a coauthor (about as surprising as … Continue reading
Running correlations – running into problems
Running correlations are a useful technique to explore how correlations between two variables vary in time or space. The correlation is calculated in a window of the first n observations, then the window is moved by one position, and the … Continue reading
Posted in EDA, Peer reviewed literature, R
13 Comments
Colour coding points in a graph
Sometimes it is useful to colour code points on a graph according to a categorical variable. There are, as always, several ways to do this. It would be possible to use nested ifelse() statements, but that way lies insanity if … Continue reading