Wednesday, 18 February 2015

NSB has a 3D printer......

A few months ago NSB and two chums called Tom and Jon bought a second hand 3D printer for a few hundred quid....

The 3D printer what we bought

Since then, Stuff . Has . Been . Made !!!!

Scale Model of Planets
The Car what No3 Son drew
A bracelet
A ruler
Mt Snowden, Mt Everest and a chocolate Mt Fuji

Scale Model of the Planets
NSB thought it would be nice to make some hemispherical scale models of the planets in the solar system.

The first attempt, at a very small scale, didn't work very well. In particular, the printer wasn't keen on flitting from planet to planet as it laid down the layers. Oh yes, and it seemed that a wall thickness of 1mm was too thin.

First bash at making scale model of planets. Not a success!

Second attempt, at a larger scale of 1: 5x10^10 and with 2mm wall thickness worked better, but printer still unhappy about moving between the planets.

Second bash at planets, still not a success

Third attempt, again at 1:50bn scale, but with just Jupiter, worked just dandy, so perhaps need to print out the planets individually. At the time of writing, this is the next job in the pipeline. In the meantime, how many people can say they have Jupiter in their kitchen drawer!

Jupiter, obviously

The Car what No3 Son drew
Something else that NSB thought might be fun was to turn a child's drawing into a real object. For this, NSB needed the help of No3 Son...

The car what No3 Son drew

The resulting 3D model

Another view

A Bracelet
A relatively quick project was this bracelet for NSB's niece, inscribed with her name.

A bracelet for NSB's niece

A Ruler
Right from the outset, NSB thought it would be cool to make some rulers that measured in mm on one side, and then in some bonkers unit (such as micro-furlongs) on the other edge. A test section revealed that rulers aren't really playing to the printers strengths and the indented markings were not actually that visible. Better off just printing out a sticker...

3D printed rulers. Not really playing to the technologies strengths...

Mt Snowdon, Mt Everest and a chocolate Mt Fuji
Continuing the "scale" theme, NSB spent about one a half weekends carefully inputting contour data for Mt Snowdon. Fuji and Everest into CAD models and then using this to create 3D printable solid structures at the 1:100,000 scale (including height above sea level)

It was surprising to see just how much bigger, and particularly taller, Mt Everest was when compared to Mt Snowdon.

Mount Snowdon (left) and Mount Everest (right), both at 1:100,000 scale

Scale is correct in height axis as well,
with base of both being at sea level

Another view of Mount Everest.

Whereas Mt Everest and Mt Snowdon were constructed with "stepped" slopes than accurately followed the contour lines, the Mt Fuji model used a draped surface approach which didn't really do as good a job in picking up the surface details and the sharpness of ridges etc. (NB: 2cm deep base to get peak to right point above sea level not shown in this image.
Mount Fuji, 1:100,000 scale

Tom thought it would be a nice idea to invert the Mt Fuji CAD drawing so that it was now a concave mould, into which he poured chocolate to make a chocolate Mt Fuji, complete with snowy peak! Genius!
A chocolate Mount Fuji and mould

Wednesday, 31 December 2014

Error Bars and how they can be improved

Interesting paper entitled "Error Bars Considered Harmful: Exploring Alternate Encodings for Mean and Error" by Michael Correll and Michael Gleicher, both from the Department of Computer Sciences, University of Wisconsin-Madison, USA.

Essentially, their work tackles the public perception of error bars, noting that bar charts using conventional error bars suffer from two main problems :
Within-the-bar bias: the glyph of a bar provides a false metaphor of containment, where values within the bar are seen as likelier than values outside the bar.

Binary interpretation: values are within the margins of error, or they are not. This makes it difficult for viewers to confidently make detailed inferences about outcomes, and also makes viewers overestimate effect sizes in comparisons.

Their proposed solution is that a more nuanced way of representing error probability is used, summarised in this image from the paper:

Ways of representing probability of error

It's all interesting stuff, and it is great to read that the public comes as being able to robustly interpret data - so long as the data meets them half way!