Those clever and inventive people at MySociety have created another slick website – now you can rate each square kilometre of the UK to help build up a map of the country’s prettiness.

See here to start voting! You’ll see a near-randomly selected photograph of a place in the UK – click 1-10 above the photo and you are on to the next one.

I hope that the Scottish Highlands come out very well – they certainly should do, many parts are very pretty when it’s not raining there…

I must declare a personal interest, as some of the photographs out there are mine. The scores for mine are pretty middle-of-the-road, as when I joined the Geograph project (which is where the photos are coming from) all the scenic areas near me had already had been well photographed.


Testing the OpenStreetMap WordPress Plugin

[osm_map lat=”51.525″ long=”-0.134″ zoom=”17″ type=”Mapnik” width=”500″ height=”300″ marker=”51.525,-0.134″]The Pearson Building, UCL Geography.

Very nice. Hopefully there will soon be an easy way to put a pin on the map using this plugin. [Update: It’s been added – thanks!]

Data Graphics

A Collection of Poor Data Graphics

This BBC article on the budget contains no less than six data graphics – and there’s something wrong with every single one.

By “something wrong”, I mean either:

  • I have to concentrate on the graphic, rather than just glance at it, to understand what it is trying to show, or
  • The numbers are distorted by the graphic – the worse kind of “wrong” as a glance at it could mislead.

The issues are:

  1. UK Budget Deficits: Apart from the unwieldy x-axis labels, showing every second fiscal year, my main gripe is the projected section of this stacked bar chart. It only works because the three projections don’t “swap over” their values at any point. But I still had to look at it for longer than necessary, to realise that the “upper value” stacked bars run “behind” the lower ones.
  2. Long-Term UK Government Debt: The use of a line chart, with smoothly flowing lines, rather than bars suggests that there are values available on a more frequent basis than every year – or that the joins between each yearly point are just artistic and so misleading. If the former, then having the unwieldy “fiscal year” x-axis, with ticks every five years, is unnecessary – why not just shift the tick marks back by 4 months and have normal years? This would be considerably easier to read. If the latter, then that’s just plain misleading!
  3. Treasury Growth Forecasts: The worst one of all. The addition of direction arrows above the positive bars (or below the negative bars) – with the value between them and the bar, and the arrows coloured the same, made me assume the bar ran up to the top of the arrow – massively increasing the value of the 2010 independent forecast, for instance. Not sure why the colours needed to change from the first chart, either, seeing as at least two of the categories have the same source in both charts.
  4. UK Claimant Count: This is a simple sequence of choropleths and as such really shouldn’t be a Flash-based chart – this is trivial to do in HTML/CSS alone, never mind Javascript. The colour sequence is odd too – a series of blues suggesting a value-based ordering, which then arbitrarily switches to purple for the final one/two bins. (The legend changing slightly for the last two!) The choropleth is also too small, so show the needed detail.
  5. Government Spending/Taxes: 3D pie-charts, tut tut! The tilt exaggerates the values at the front, making them seem bigger than they are.
  6. UK Rescue Plans: The circles are correctly scaled in 2D rather than 1D – a common mistake averted. However, they unnecessarily overlap with each other, so partly obscuring the genuine ratios

The Beeb designers need to take a read of Tufte and not go down the Microsoft Excel route!


MSc Dissertation

Last year (2007-8) I studied for an MSc in Geographic Information Systems, at City University London. The course was taught by an excellent team of academics and I can thoroughly recommend it as a good, technology-focused introduction to GIS. The highlight was the field-trip, a week away in the Lake District, carrying out three two-day projects, each involving planning, data gathering, preparing and presenting the findings.

The summer last year was spent researching and writing the dissertation. It is entitled “Use of a GIS for Production and Maintenance of Street Orienteering Maps: Can a GIS and Spatially Aware Data add Value to Orienteering?” and can be downloaded from here (24MB, 102 pages).

You almost certainly don’t want to read 102 pages, so there is an extended abstract here (1MB, 5 pages), entitled “Creating and Maintaining Street Orienteering Maps using OpenStreetMap”, which appears in the “Proceedings of the GIS Research UK 17th Annual Conference”, aka GISRUK 2009. I presented a poster summarising the work at the conference, which is reproduced below – linked to a larger version.

Poster for GISRUK 2009