Last Resort Software

Some time ago I wrote an R function to fit an ellipse to point data, using an algorithm developed by Radim Halíř and Jan Flusser 1 in Matlab, and posted it to the r-help list . The implementation was a bit hacky, returning odd results for some data. A couple of days ago, an email arrived from John Minter asking for a pointer to the original code. I replied with a link and mentioned that I'd be interested to know if John made any improvements to the code. About ten minutes later, John emailed again with a much improved version ! Not only is it more reliable, but also more efficient. So with many thanks to John, here is the improved code: fit.ellipse 0)] f Next here is a utility function which takes a fitted ellipse and returns a matrix of vertices for plotting: get.ellipse And finally, some demo code from John: create.test.ellipse 1 Halíř R., Flusser J.: Numerically stable direct least squares fitting of ellipses. In: Proceedings of the 6th International Confere...

A recent upgrade of NetBeans from version 7.1 to 7.2 seemed to cause problems with Maven multi-module projects. Normally, a newly created class in one Maven project module will be instantly visible to other modules open in the IDE, at least for editing purposes. But after the upgrade I started to see "cannot find symbol" errors in the IDE for such new classes. Despite this, a full build of the project would succeed. Googling for an answer drew a blank. The usual kludge fixes such as deleting the NetBeans cache (for NetBeans 7.2 on OSX this seems to have moved from ~/.netbeans to ~/Library/Application Support/Netbeans) and re-indexing my local repo didn't make any difference. Despair loomed. Happily, I stumbled across the cause of the problem after only a moderate amount of pulling of hair and gnashing of teeth. My Maven settings.xml file included this section... <profiles> <profile> <id>netbeans-ide</id> <...

I've been working on a small Java application with a colleague to simulate animal movements and look at the efficiency of different survey methods. It uses the GeoTools library to support map projections and shapefile output. GeoTools is great but comes at a cost in terms of size: the jar for our little application alone is less than 50kb but bundling it with GeoTools and its dependencies blows that out to 20Mb. The application code has been changing on a daily basis as we explore ideas, add features and fix bugs. Working with my colleague at a distance, over a fairly feeble internet connection, I wanted to package the static libraries and the volatile application into separate jars so that he only needed to download the former once (another option would have been for my colleague to set up a local Maven repository but for various reasons this was impractical). A slight complication with bundling GeoTools modules into a single jar (aka uber-jar) is that individual modules make ext...

For a current project I needed a polygon smoothing algorithm in Java that: generated a curve passing through all polygon vertices had some way of controlling tightness of fit to the polygon fell within the domain of my very limited mathematical grasp Luckily for me, Maxim Shemanarev had not only worked out a Bezier curve algorithm that satisfied all of the above, but he also published a beautifully clear explanation of how it works note 1 . Maxim also provided some illustrative C code, but my Java implementation below is an independent effort (ie. all mistakes are my fault). It uses geometry objects from the JTS library . Below is an example of the algorithm in action. The original polygon is in black. The smoothed polygon composed of fitted cubic Bezier segments is in blue. The red points and lines indicate positions of the Bezier control points calculated with Maxim's algorithm. There are a pair of control points for each vertex of the input polygon. The tightness of fit of th...

This is another article on the theme of defining groups in a hierarchical classification. A previous article described homogeneity analysis to visualize how any well any number of groups, defined at the same level accounts for the variability in the dataset, as measured by within-group pairwise distances. Here we will look at testing whether splitting a particular group into two daughter groups is supported by the data. Suppose we have a group G containing N objects which we are considering splitting into two groups G1 and G2 consisting of N1 and N2 objects respectively. A Monte Carlo approach to testing whether this is worth doing is as follows: 1. Calculate the mean within-group pairwise distance that would result from defining G1 and G2. 2. Randomly allocate objects to two groups of sizes N1 and N2 and calculate the resulting mean within-group pairwise distance. 3. Repeat step 2 a large number of times (e.g. 999) to create a vector of Monte Carlo means. Add the actual mean from ste...

The proto package is my latest favourite R goodie. It brings prototype-based programming to the R language - a style of programming that lets you do many of the things you can do with classes, but with a lot less up-front work. Louis Kates and Thomas Petzoldt provide an excellent introduction to using proto in the package vignette . As a learning exercise I concocted the example below involving Bayesian logistic regression. It was inspired by an article on Matt Shotwell's blog about using R environments rather than lists to store the state of a Markov Chain Monte Carlo sampler. Here I use proto to create a parent class-like object (or trait in proto-ese) to contain the regression functions and create child objects to hold both data and results for individual analyses. First here's an example session... # Make up some data with a continuous predictor and binary response nrec x y # Predictor matrix with a col of 1s for intercept pred colnames(pred) # Load the proto package li...

I've spent more years than I care to remember analysing vegetation survey data (typically species abundances in plots) using a variety of software including my own algorithms coded in FORTRAN and C++. A recent query on the r-help list, about how to determine the number of groups to define in a hierarchical classification produced with the hclust function, prompted me to unearth one of these algorithms, homogeneity analysis 1 , which can help to visualize how different levels of grouping partition the variability in a distance matrix. This algorithm is extremely simple. The classification is progressively divided into groups, with all groups being defined at the same dendrogram height. At each level of grouping, the average of within-group pairwise distances is calculated. Homogeneity is then defined as: H = 1 - Dav within-group - Dav total where Dav total is the average pairwise distance in the dataset as a whole. For data were there exist well-defined clusters of values, a...