Everything is Connected

Prezi! And amusing facial hair regulations

2010-12-04T13:23:00.000-08:00

I saw a presentation by DECC's youth panel today, which used Prezi (pronounced "pre-zee"). Prezi is software available for-free or for-money at prezi.com. Inspired by what I saw, I registered and wrote my own Prezi and it is here. The presentation includes guidance on permissible facial hair at Chevron oil refinery in Pembroke.

Chevron Refinery 2010 on Prezi

learning to use Gnome in ubuntu

2010-06-21T08:01:00.000-07:00

I'm a crusty old unix user, and I have been using nice old twm as my window-manager for ages (since about 1991). Everything worked nicely through several generations of linux, but eventually last month it was just too much hassle to get things like wireless and ethernet to work, so I gave in and switched to gnome. (Philosophically, it seems a broken arrangement that has wireless and ethernet enabled only when the X-server is working... what if the Xserver is down, and you want to connect to the outside world to install a patch?! What if you want remote users to be able to log in to the machine?)
Anyway, on this page I will note down a few of the niggles I had and how I resolved them.
Niggle 1: which is the window-manager, and why, and how are its settings being configured?

I wanted to restore behaviour to be like my twm set-up; I had a very confusing time, because it seemed along the way I had changed window manager from compiz to metacity (which meant none of the compiz settings were having any effect any more). Where was this choice determined? The answer was that I had at some point switched the Appearance setting to have "no animations", instead of "some exciting animations, oo oo ooh", or "lots of stupid 3-dimensional swishy rubbish", and this silently switches window manager! In order to get the other useful features of compiz (Eg suppressing some window titles), you have to switch back from "no animations" to one of the stupid animation levels.
Then the question is how to switch off all the animations one by one...
Niggle two: window-switching
I used to have my windows on the desktop in a cycle. Then I could use one key to rotate up the cycle, and one to rotate down it. With the default "Alt-tab" way of doing things, the windows get reordered every time you switch windows. This doesn't suit my brain's way of navigating. What's the compiz way of getting a cycle? I haven't figured this out yet.
I have a partial solution...
Under compiz settings -> window management -> application switcher, I have selected under "Bindings"
Next window = Alt-tab
Previous window = Ctrl-tab
Next window (no popup) = F10
Prev window (no popup) = F12
and under "General"
Speed 19
Timestep 0.1
Saturation 100
Brightness 100
Opacity 100
Bring to front NO (but this does not seem to work, see below)
Zoom 0
...So Now I can rattle through the windows in one direction with F10 and in the other with F12... except, F10 doesn't leave the order of the windows unchanged. Is this a bug? So actually F10 simply exchanges two windows only.

Add word to dictionary (fix instructions)

2010-06-04T08:24:00.001-07:00

In my new job I have been forced to use Microsoft Office for the first time.
One of the minor painful issues has been the impossibility of adding "misspelled" words to the custom dictionary so that the "correct spelling?" dialogue doesn't keep nagging. The "add word to dictionary" option was always greyed out, and I couldn't figure out how to get this option to be enabled.
At last, I have been given a fix, and it works.
Custom Dictionary fix instructions

1. Ensure that any Office 2007 products that are currently open are closed before proceeding.
2. Find your custom dictionary located under P:\Office97\Custom.dic, create a backup by right clicking on the file and selecting copy. Paste the file within the same folder, a backup will be created called “Copy of Custom.dic”.
3. Now open Custom.dic file by double clicking on it. The file will open within notepad.
4. From the File menu select “Save As”. In the “Encoding” field set the encoding type to be Unicode.
5. Select Save, click Yes to confirm that the file should be replaced.
6. Close notepad.
7. Start Microsoft Word and confirm that you can add words to the dictionary.

Happiness!!!

'Artificial Shakespeare' breakthrough announced by scientists

2010-05-21T00:26:00.000-07:00

Scientists in the US have succeeded in developing the first synthetic Shakespeare.

The researchers constructed a Shakespeare play by assembling "sentences" using 26 sterile letters then printing it out on an old printer.

The Scientists at Craig Venter laboratories emphasize the massive potential of their technology. "We copied 'Romeo and Juliet' a few words at a time and stiched them all together, and ended up with a copy of the original play. We thus have made synthetic William Shakespeare, and will be able to transform literature."

Critics have asserted however that this work, albeit impressive, should not be described as producing "artificial Shakespeare". Speaking to the Gullible Broadcasting Corporation, Nobel
Laureate Sir Paul Nurse said "You lot shouldn't be so gullible. What they have actually done is copy Romeo and Juliet a few words at a time and stiched them all together."

Other news: playing God?

Bioethicists have reacted with caution to the announcement that scientists in the US have transplanted blood from one human to another.

Electoral systems

2010-05-09T10:06:00.001-07:00

I've been reading up on voting systems and "proportional representation". The BBC has a nice article on what other countries do. If you are interested in a bit more detail about voting systems from an academic who has thought about them a lot, see Denis Mollison's article about STV and his paper for publication in the RSS's journal.
Denis, a statistician at Heriot-Watt University, recommends the STV system with multi-candidate constituencies. Its implementation would require a computer.
[If you're interested in more exotic methods that require not only a computer but also the use of randomness, then please see Probabilistic electoral methods, representative probability, and maximum entropy by Roger Sewell, David MacKay, Iain McLean.]

How to make an animated gif with small file-size

2010-02-08T09:42:00.000-08:00

I wanted to make an animation showing the evolution of global temperature and the number of sunspots over the last 129 years. I used gnuplot to spit out 129 png files (each of size 2k to 14k), then the question was how to make the animation.
Initially I tried:
convert -delay 10 -loop 0 *.png -delay 1000 2009.png Anim.gif
which made a grand-looking animated gif, but its size was enormous - over 1 megabyte! Many thanks to Patrick Welche for telling me to try the deconstruct flag:
convert -delay 10 *.png -deconstruct Anim.gif
which makes a 161k animation.
Here is a link to the resulting animated gif (161k)

Bayes' theorem makes it into the BBC

2009-07-18T03:29:00.000-07:00

This article about false positives seems quite well done.

The BBC article links to an animation on the Understanding uncertainty website. I'm not sure if the animation is a good teaching aid - it shows the test being applied with a great random rattling of the faces of the 100 humans. I think it would be better to reveal the truth early so the viewer can understand the generative process, instead of it looking like every individual is being utterly randomized. I also find it confusing that colour (Red/green) is used initially to encode test result, then it is later used to encode the underlying truth. I think two different dimensions should have been used.
I'd reveal the truth, show the generative process, then hide the underlying truth to emulate "what the tester sees"; then reveal the truth again.

ps... I find it a bit uneasy-making that the whole set-up involves "detecting terrorists". The implicit endorsement of "war on terror"; the endorsement of the campaign to "be suspicious of your neighbours, look for stuff in their bins, and get them arrested and interrogated".

How to handle copyright forms from dinosaur journal-publishers

2009-06-28T03:11:00.000-07:00

If you want to publish an article with a dinosaur journal that says "you can't publish your academic work on the internet any more; all your copyright are belong to us", here is the recommendation: fill in the
SPARC addendum and attach it to their copyright form.
Assuming the journal publisher accepts this addendum, your work can then benefit from the modern day miracle of efficient communication.

One hundred and eighty!

2009-06-03T12:14:00.000-07:00

Three students in the inference group had their PhD vivas on June 1st, 2009.
Congratulations to Ollie Stegle, Ryan Adams, and Keith Vertanen on superb PhDs!

Star Wars - Retold

2009-02-15T05:47:00.000-08:00

Star Wars - Retold (by someone who hasn’t seen it)

Star Wars: Retold (by someone who hasn't seen it) from Joe Nicolosi on Vimeo.

Hurray for the internet - I haven't laughed that much for a while!

only self-teaching has any lasting value

2009-02-08T15:50:00.000-08:00

"only self-teaching has any lasting value" is a quote from this essay by John Gatto,
"Why Schools Don't Educate".

My own thoughts on this topic are in (1) a short essay on exams - "everyone should get an A" - which notes how exams interrupt education; and (2) a one-page suggestion on "how to teach".

Gaussian Processes are ill-conditioned (Example)

2009-01-31T07:46:00.000-08:00

Someone asked me "why did you say in the Gaussian processes tutorial that Gaussian processes are ill-conditioned?" Here is an answer in the form of a tiny exercise.

Take the simplest case where the function is just a constant.

y(x) = Y for all x. (Y is Gaussian distributed and not known a priori)

Make a GP with the appropriate covariance function.

Now

(1) think about the inference of the GP given data {x,t}. (Not very difficult is it?)

(2) see what happens when you use the standard GP matrix inversion formalism to solve the problem. What is the well-conditioned-ness of the matrix you must invert?

Soaring, Cryptography and Nuclear Weapons

2009-01-09T14:52:00.001-08:00

Please go to nuclearrisk.org and read renowned cryptographer Martin Hellman's article Soaring, Cryptography and Nuclear Weapons.

If you feel the urge to pass the message on, pass it on.

Why traders make bad investments

2009-01-03T09:46:00.000-08:00

I heard this on Radio 4 - Paul Wilmott's explanation of why traders will rationally put their bank's money into suboptimal risky investments.

(Of course this isn't "the" reason why bad investments happen. Rather, this cartoon presents a simple world in which we can see why traditional incentives for employees lead to bad outcomes.)

Imagine Joe is a trader at a fancy investment bank where all the many other traders are lemmings who unanimously invest over the next 6 months in deal A, which has a 50% chance of making a big return, and a 50% chance of not. Joe is a wise trader and has identified deal B, which has a 75% chance of making a big return for his bank. Deals A and B are independent random variables. What should Joe do with the bank's money?

Well, we need to specify Joe's utility function, which is dominated by his bonus. The rule at his bank is that Joe gets a huge bonus if the bank makes a big return, AND Joe's chosen investment made a big return. Otherwise, no bonus.

If Joe invests in deal A (like all the other lemmings alongside whom he works) then he has a 50% chance of getting a huge bonus.

If Joe instead invests his relatively small part of the bank's money in deal B, then he will get a bonus only if _both_ deal A and deal B come out as successes - because the bank will get a big profit only if deal A succeeds. So Joe has a 37.5% chance of getting a big bonus.

So Joe's rational decision is to invest the bank's money in deal A, the inferior investment.

A glimpse of Dubai

2008-11-18T02:09:00.000-08:00

I spent three days in Dubai for a World Economic Forum "global agenda council" gathering. We were in the Jumeira Hotel complex. My favourite sight of the trip was this abandoned car sitting outside the hotel. 'My other abandoned cars are a Porsche, a Hummer, and a Rolls', perhaps?

Typo hunting

2008-11-15T14:28:00.001-08:00

I finished the book Sustainable Energy - without the hot air a week ago, and got the proofs today. I promptly managed to find a couple of small typos that I hadn't spotted before. Nothing too bad, happily. Unlike the English-speaking makers of this road sign in Swansea...
The Welsh translation translates back as "I am not in the office at the moment. Please send any work to be translated."

How to use latex2eps.sh to make beautiful postscript fragments

2008-08-28T12:34:00.000-07:00

latex2eps.sh
is a super utility I picked up when I started using magicpoint.

Simple example

If file1 contains


$$\mbox{Power} = \frac{1}{2} \rho A v^3$$

and you run


latex2eps.sh Wind0 < file1

then you will get a file Wind0.eps, which is a beautiful postscript file looking like this

More complex example: Package options

You can give package options when running latex2eps.sh. For example, if you like \usepackage{booktabs} or \usepackage{colordvi}, here is what you can do.
Let's say that file2 contains:


\begin{tabular}{lcl}\toprule
\multicolumn{3}{c}{ \ \OliveGreen{$v=6$\,m/s} (force 4)}\\ \midrule
Wind farm  &  & \Blue{2\,W$\!$/m$^2$ flat ground} \\
 \bottomrule
\end{tabular}

Then we run the command:


latex2eps.sh -sty colordvi -sty booktabs Wind1 < file2

and we get this lovely object:

My favourite blogs

2008-08-28T12:29:00.000-07:00

The blogs I read are:

Radford Neal's blog (Radford's in CS and statistics at Toronto)
Michael Mitzenmacher's blog (Michael's in Engineering at Harvard)
Jonathan Yedidia's blog (Jonathan is at MERL in Boston) (though he seems to have not posted anything for a long time)

The blogs I write are:

This one

Sustainable Energy - without the hot air

Identity Theft

2008-08-28T10:16:00.000-07:00

This morning I got a letter in the post from Virgin mobile. It said "Thank you for arranging to pay your mobile bills by direct debit; please contact us if any of these details are wrong".
I didn't recall arranging to pay for a Virgin mobile phone by direct debit, so I phoned the number. After about 30 minutes of listening to annoying music as the Virgin organization tried to figure out what to do with my phone-call - it wasn't appropriate to talk to the direct debit department about it, of course; no, I needed customer service... - I eventually spoke to a nice lady whom I'll call Mary. Mary asked me a bunch of questions about the letter and me (including, strangely, asking for my Virgin mobile password, which didn't seem relevant), and then she explained (after another few minutes of elevator music) that it seems someone else is using my identity (name, address, date of birth) to register a phone. It's not clear to me why they would do this, as the direct debit wasn't coming out of my bank account. Anyway, she said she would file a Fraud Alert with the Credit Report people, so maybe my name will be blacklisted from credit requests in the future.
She recommended that I contact Equifax to find out if other credit application frauds have happened using my name, and whether my credit score has been damaged. So I took at look at the Equifax webpage, and found a bunch of Equifax organizations, all with the same logo, all rather keen to take my name, address, date of birth, credit card number, and other security details. And keen to charge me money to answer the question "what is my credit history, and is anything funny going on?" I'd never heard of Equifax before and once bitten.... I feel rather reluctant to go typing all my personal details into a form. It seems Equifax are an international organization that handles everyone's credit history. And then charge everyone in sight for that information, including banks and the innocent civilian.

All about this blog (ITILA)

2008-08-28T09:57:00.000-07:00

The ITILA blog is named after Information Theory, Inference, and Learning Algorithms, the first book I wrote. (It's available free online, so do help yourself!) One of the themes of that book was that everything is connected. Not only information theory and machine learning, which are two sides of a single coin; but also communication, data compression, evolution, sex, satellites, discdrives, solitons, thermodynamics - pretty much any topics you care to mention. ITILA is a textbook about information theory whose goals include bringing out these connections, and making information theory and statistics fun.
This blog is not only for posts about ITILA the book.
I'm going to use this blog to post work-related stuff - all the things that I used to slap into my now rather unwieldy home-page;
and for occasional personal items that I think might interest work colleagues.
Anything I'm excited about, I'll probably put a post here.
By the way, I do have another blog, withouthotair, which is where I post anything to do with energy. Like this blog, that blog is also named after a book - "Sustainable Energy - without the hot air" - which should be published in November 2008. (It's also free online.)