I’ve just finished the very short introduction to statistics, and now I get it!

So the X axis is something, like cleanliness or laziness, and the Y axis is how many people are that clean or lazy.

If the distribution looks like this, then if you are in the middle, then you are more normal than most!

It is all in full swing, but here are some amazing pictures from SG this year!

the picassa albums

Rudi Stouffs (check out some of his work here) videoed some of the evening presentations.

There are some great talks in this bunch, Sam Joyce’s is particularly worth a go for a bit of a reflective view on structural engineering!

Mine is a dramatic speed up of the talk I gave at UTS last year. If you are interested, the slides are here.

I got a bit sleepy in a meeting the other day and I started drawing. It was one of those darkened room affairs with an impeding food coma, so i needed something to ease the pain.

It just so happened that I was drawing with my Livescribe pen. I’m going to post this now, but I’ll finish it off later.

Here’s a photo of the pen scribble in my book. It’s about 50mm across in real life.

The finished version is missing a lot of the feinter lines as the pressure sensor on the pen isn’t as responsive as something like a good wacom tablet.

Click for another pattern.

Alasdair Turner has posted this processing version of Paul Coates’ emergent Voronoi algorithm from his book Programming Architecture

Alasdair says “This is an implementation of an emergent Voronoi diagram, following an algorithm presented in programming.architecture by Paul Coates. It adapts the algorithm slightly to give a good convergence. Coates simply says “move away from the nearest node”, which is wonderfully elegant!“

It is an impressive and clear way of explainign what is really going on behind all those funny shapes people make.

I have a curious obsession with scalpel blades, there are so many types, and the Swann Morton website is so bad, that I really didn’t know what they all looked like.

They are handy for good model making, but people only ever seem to use 6, 10 and 10A. I’m pretty sure that there could be a good use for the 9, the 12D the 16 and the 40 in any studio!! (Just think about how beautiful a scalpel rack would be.)

This is just the few that fit into a number 3 handle. There are loads more that are specific to very fine dental work etc. that look really handy, but this is a start!

No. 06	No. 09	No. 10
No. 10A	No. 11	No. 11P
No. 12	No. 12D	No. 13
No. 14	No. 15	No. 15A
No. 15C	No. 15T	No. 16
No. 40	No. E/11	Sabre D/15
Sabre E/11

Just to allay any fears, I am not feeling suicidal and wanting to find the right tool to top myself in the most elegant way (probably a 40)

This is really very sad, but I like it. It goes along with my doctrine of extreme laziness.

For the uninitiated, the idea is that you don’t specify the formatting for each bit of text, you specify a style for it, then you can go back and change the style and all applications of that style change throughout the document. you can even nest styles, but I’ll spare you that.

Anyway, you can do this in InDesign, Word, Open Office Writer etc. and it is great. The problem occurs when you need to intercept a document that someone else has done a blinding job of cocking up.
In InDesign it is easy, there is a button to clear style overrides, but in word [ctrl]+[space] clears the local formatting, and [ctrl]+[q] clears local spacing overrides. I got this tip off Susan Daffron at logicaltips.com

This is a mega dorky topic, but really, if you use styles you’ll have much more time to invest in snorting cocaine off strippers and going base jumping.

…yup

The task I set was to make the dude that Daniel Shiffman draws in his example on the learning processing page stick to the mouse, and scale with the mouse’s position.

Not only that, but to do it in a readable way with minimal use of explicit values so that we can use it to make an object for agent simulations next week.

This isn’t easy, so use a lot of sketches, and diagrms on graph paper to help you. A programmers best tool is their brain, and then second best is a notebook!

My example of one way to do it is after the fold. I don’t expect you to solve this all on your own (but extra brownie points if you do) so go through my code and get a handle on it.

This is an example of one way (there are many) that you could have written your code to allow for the dude to be drawn

void setup()
{
  size(400,400);
  rectMode(CENTER);
}
void draw()
{
  //redraw the background
  background(255,255,255);
  //expose some variables
  int centreLineX = mouseX;
  int bodyCentreY = mouseY;

  //********some internal variables***********
  int headCentreY = bodyCentreY-30;
  int bodyHeight = 100;
  int bodyWidth = 20;
  int headSize = bodyWidth*3;
  int eyeWidth = (headSize/3)-2;
  int legSize = 20;

  //**********draw the dude***************
  //body
  rect(centreLineX, //centroid X
       bodyCentreY, //centroid U
       bodyWidth,
       bodyHeight);
  //head
  ellipse(centreLineX,
          headCentreY,
          headSize,
          headSize);
  //left eye
  ellipse(centreLineX-19,
          headCentreY,
          eyeWidth,
          headSize/2);
  //right eye
  ellipse(centreLineX+19,
          headCentreY,
          eyeWidth,
          headSize/2);
  //left leg
  line(centreLineX - (bodyWidth/2),            //startX
       bodyCentreY + (bodyHeight/2),           //startY
       centreLineX - (bodyWidth/2) - legSize,  //endX
       bodyCentreY + (bodyHeight/2) + legSize);//endY
  //right leg
  line(centreLineX + (bodyWidth/2),
       bodyCentreY + (bodyHeight/2),
       centreLineX + (bodyWidth/2) + legSize,
       bodyCentreY + (bodyHeight/2) + legSize);
}

I’ve delibearly left out the scaling to keep the code readable in the version above. Here it is with the scaling involved.

void setup()
{
  size(800,800);
  rectMode(CENTER);
}
void draw()
{
  //redraw the background
  background(255,255,255);
  //expose some variables
  int centreLineX = mouseX;
  int bodyCentreY = mouseY;
  /* This is the scaling part
   * it is a bit tricky to figure out, but think about
   * it as a diagram on some graph paper. The mouseX and
   * mouseY numbers get extremetly large, not very far
   * away from the origin, so they need to be scaled down
   * a lot or our man becomes a tiny dot.
   *
   * The +1 is so that when mouseX or mouseY = 0 that we
   * don't get a 'divide by zero' error*/
  int scaleX = mouseX/100 + 1;
  int scaleY = mouseY/100 + 1;

  //some internal variables
  int headCentreY = bodyCentreY - (30/scaleY);
  int bodyHeight = 100;
  int bodyWidth = 20;
  int headSize = bodyWidth * 3;
  int eyeWidth = (headSize/3)-2;
  int legSize = 20;

  //draw the dude
  //body
  rect(centreLineX, //centroid X
       bodyCentreY, //centroid U
       bodyWidth / scaleX,
       bodyHeight / scaleY);
  //head
  ellipse(centreLineX,
          headCentreY,
          headSize / scaleX,
          headSize / scaleY);
  //left eye
  ellipse(centreLineX - (19 / scaleX),
          headCentreY,
          eyeWidth / scaleX,
          headSize / 2 / scaleY);
  //right eye
  ellipse(centreLineX + (19 / scaleX),
          headCentreY,
          eyeWidth / scaleX,
          headSize / 2 / scaleY);
  //left leg
  line(centreLineX - (bodyWidth  / 2 / scaleX), //startX
       bodyCentreY + (bodyHeight / 2 / scaleY), //startY
       centreLineX - (bodyWidth  / 2 / scaleX) - legSize / scaleX,  //endX
       bodyCentreY + (bodyHeight / 2 / scaleY) + legSize / scaleY);//endY
  //right leg
  line(centreLineX + (bodyWidth  / 2 / scaleX), //startX
       bodyCentreY + (bodyHeight / 2 / scaleY), //startY
       centreLineX + (bodyWidth  / 2 / scaleX) + legSize / scaleX,  //endX
       bodyCentreY + (bodyHeight / 2 / scaleY) + legSize / scaleY);//endY
}

If you’ve got this far, you’ve done very well as this is some tricky stuff!

The SmartGeometry Group is now accepting applications for…

Open Platform

A 3-day Workshop at TU Delft University 27th-29th November

This event will complement SmartGeometry’s annual workshop and conference event, bringing many of SmartGeometry’s experienced tutors to guide participants in exploring computational design through a variety of technologies.

Click here to view the Event Poster

The event will run in two parallel streams: one stream will feature training in Generative Components (GC) and Grasshopper, while the second will be a workshop where participants have the opportunity to propose and work on individual design projects using a variety of software platforms. Both streams will offer participants close interaction with tutors from practices and universities known for their accomplishments in computational design and analysis, and complex geometry. Additionally, the stream featuring GC and Grasshopper will feature tutors from the companies behind each platform, Bentley and McNeel. Those working on individual projects will need to be proficient in their chosen platform but can draw on tutor’s skills in various technologies — for instance Processing, C++/OpenGL, RhinoScript — as well as tutors’ expertise in complex problems of geometry, structure, algorithms, and analysis. Participants will be able to change streams as it suits the development of their work.

In the evenings, participants can attend presentations demonstrating the latest projects and research in computational architecture and engineering.

Fees for participation are as follows:

Students/Academics = 75 euros

Professionals = 200 euros

Fees cover training and/or workshop room, and evening presentations.

To register send an email to applications@smartgeometry.org with the following information:

Name

Email Address

Physical Address

Telephone number

Affiliation

Are you a student or professional?

Order of preference between: Grasshopper Training, GenerativeComponents Training, or Open Workshop. If you can/want to only attend one in particular, please list only that option.

The deadline for applications is November 15th.

All applicants will be notified on their acceptance on November 18th and to arrange fee payment.

Questions can be directed to applications@smartgeometry.org