A dream of sleep -

Did anyone catch Horizon on Tuesday?

There was an interesting psychiatric test. Imagine you have two jar, one containing 80 red balls and 20 blue balls, the other containing 20 red balls and 80 blue balls. In the test you are presented with balls having been taken at random from one of the two jars* [i.e. if its the majority blue jar, balls will continue to be taken from this jar, with each being selected at random], and apparently the point at which you make your decision on which jar the balls are from is indicative of some mental conditions.

Now, suppose this is a fair test, and lets simplfy by imagining an infinite number of balls (so we don't have to deal with changing probabilities), as a friends list of mainly scientists, engineers and mathematicians, at what point should we be confident in our choice of jar?

( Bayesian Stuff )

* Yes, I watched the show, but some people didn't and I wouldn't want to spoil the test for them.

If as a citizen you would like to form well-considered views on a culturally divisive risk issue - for example, global warming, or gun control - find a knowledgeable person who shares your general cultural outlook but who disagrees with you. You are likely to give this person's arguments a sympathetic hearing, which will help offset the natural disposition we all have to dismiss as unreliable and biased the arguments of persons whose basic outlooks are different from our own.

Dan Kahan, Yale Law School

( Another, not so good quote. )

( Iron Man: Review )

Predictions are fun!

Make predictions for three events which will happen within the next: year, decade and century. Specify whether the event would have been true were you to make the prediction this time previously - i.e. for a prediction of an event in the next year, would it have come true if you had predicted it a year ago.

You can choose what headings you want to make predictions under.

( Mine )

I need some help with regular expressions if there is anyone out there who knows enough about them:

( The Problem )

EDIT: It works, by using a different method to how I was thinking though.

COP OUT!

Complete and utter cop out.

I am not amused.

Now to watch the rest of the episode, and attempt to get off my high-horse.

EDIT: Still not happy, but, well... I want a Jenny spin-off.

In a further poll regarding my thoughts on the New Media, I would be obliged to people who filled in the following:

Edits: Okay - questions are now in place above the poll.

( Creativity , Fairness and Income )

As some of you may know I'm pretty darned interested in voting systems, and more specifically trying to ensure that a system is fair and representative.

I don't have too much work to do right now so I thought I'd have a play with implementing some system with actual data, and thus try to determine which produces the 'best' results. This is where you people come in.

We are going to be hosting an election for the 'best' colour. The candidates are:
i) Red,
ii) Blue,
iii) Green,
iv) Yellow,
v) Orange,
vi) Purple,
vii) Pink,
viii) Black,
ix) White,
x) Brown.

Can you please comment with the following [See below for descriptions of the voting methods]. Assume in all cases that a single member is to be elected:
i) A 'true' ordering of your preferences - that is a ranking independent of any tactical concerns regarding which options you think are most likely to win.
[For the following two please consider strategic voting concerns (see notes below) - if these do not change your preferences then you may indicate 'as above']
ia) A ranking of candidates for use under an 'instant run-off' (single transferable vote) voting system.
ib) A ranking of candidates for use under a 'Borda count' system.
ii) Which single candidate you would vote for in a 'plurality' (first past the post) voting system.
iii) Which multiple candidates you would vote for in an 'approval voting' system.
iv) A score for each candidate in the range of 1 - 10 points (using whole numbers only) for using in a range voting system.

If you have any questions or comments can you make independent comments for votes and queries as to ensure a private ballot comments and screened by default - however I may choose to un-screen useful queries or comments.

NOTES:
* ( Plurality voting )
* ( Approval voting )
* ( Range voting )
* ( Instant Run-off )
* ( Borda Count )

EDIT: There is also a straw poll being conducted by atreic on her journal now if you want some idea of the popularity of various options.

I've recently started going to the Gym, and I'm finding that my calf muscles are aching a lot. I don't think that there is any injury as such, but it is a little painful to walk down stairs... so - as I have some people on my flist who are vaugely medical, or at least exercise more often than me - is this:

i) Something I should be more worried about?
ii) My body telling me to quit exercising for a while?
iii) My body telling me to cut back a while?
iv) Of no concern at all - just a consequence of exercise which will go away / something I need to put up with?
v) Good - an indication that the muscles are being strongly exercised and will thus grow stronger until I become SUPER NEIL - ABLE TO LEAP LARGE PUDDLES IN A SINGLE BOUND!
vi) Something else.

Ah - Wikipedia. April 1st is when you really shine, particularly in recent years in which you have actually highlighted true, but confusing stories, for instance:

DID YOU KNOW:
...that in a few villages and towns of southern France and Spain it is illegal to die, and that there are attempts to have the same law in a town in Brazil?
...that Ben Affleck died while shoveling snow outside of his house, leaving behind an unexpectedly small estate speculated to be worth as little as US$20,000?
...that John F. Kennedy was shot dead in an ambush by government agents who had foreknowledge of his whereabouts?
...that American entrepreneur Timothy Dexter defied the popular idiom and actually made a profit when he sold coal to Newcastle?

It's a weird, but really cool ancient greece based radio comedy show. I though the following quotation was really good fun. Its between Socrates and another character I can't be bothered to find out the name of!

"So - no Mrs Socrates today then?"
"No - left her at home. Although I expect by now she's helping your brother in the food preparation area. The woman's incorrigible it pains me to say it - I mean I married her in all good faith... but between you and me you do realise... She's Straight?"
"No - Xanthippe's straight?!"
"Keep your voice down - you hadn't guessed then? Zeus they're cunning these heterosexuals."
"Its such a shock - I mean; is this a harmless lifestyle choice or..."
"Oh hardly. Not when she fancies people of the opposite gender."
"Oh!"
"Oh - she doesn't fancy me thank goodness."

[Edited to reflect comments - transcriptions is hard.]

Which Discworld Character are you like (with pics) created with QuizFarm.com

You scored as Lord Havelock Vetinari You are Lord Vetinari! Supreme ruler of Ankh-Morpork! Cool, calculated, and always in control. You graduated from the assassins guild, but failed a course on stealth and camouflage, because the professor never saw you there (even though you attended every class). You always seem to know what everyone is thinking, and after a conversation with you, people feel that they have just escaped certain death. Lord Havelock Vetinari 88% The Librarian 75% Commander Samuel Vimes 69% Carrot Ironfounderson 63% Esmerelda (Granny) Weatherwax 63% Cohen The Barbarian 56% Death 56% Gytha (Nanny) Ogg 44% Greebo 38% Rincewind 38%

I've just taken an "implicit association" test regarding my feels towards the American presidential candidates and apparently I strongly favour Hilary Clinton, favouring Obama only a little over McCain who I favour only a little over Mike Huckabee.
New Scientist discusses the test at: http://www.newscientist.com/blog/shortsharpscience/

Basically the test will ask you to identify either an image or a word as 'good' or 'bad' - good words include 'friend', 'happy', 'pleasant', bad words include 'hate', 'enemy', etc. Interspersed with the words are images of candidates which at the beginning of each round will be assigned to being good (i.e. one does not select good words, but good words or images of candidate A) - another candidate images will need to be placed with bad association.

Now I wonder whether this test may not so much have identified my attitudes towards the four candidates, but instead just shown how differentiable they are - that is it is a lot easier to identify an image of Hilary (a woman) over any of the men, and easier to identify Barak Obama over the two white men. I question whether the test truly identifies my attitudes, or simply reveals that I am better at identifying a woman over a man?

Presumably this test should have been calibrated for this, but given the length (under ten minutes) I doubt whether the effect could fully have been corrected for. Or maybe I do indeed, apparently like the author of the New Scientist blog entry, have a secret crush on Ms. Clinton.

After about 10 months my 'NAM' (National Astronomy Meeting) freebie back-pack (which I have used daily for carrying stuff to and from the Office, has been to London at least thrice, the Czech Republic once, and the Isle of Wight a good number of times) has finally broken to such an extent that it is no longer usable...

Let us all have 1 minutes silence.

Don't you just hate it when you like a band and then it turns out they're Canadian?

( Read more... )

EDIT: Am listening to them now, reliving teenage angst :: smile ::
Sings: "Yesterday was the worst day ever, and tomorrow won't be better... it's history repeating"

I'm quite chuffed - I've found a map of the UK population density on the Internet.

Its something for which I have been searching, on-and-off for a fair amount of time, and while the current map isn't perfect (the resolution is quite bad, 312 by 312 points covering a larger area than the UK itself, and I suspect that it isn't a true representation of the population density, but in fact a representation of the density of the population of the local government region within which a point is found [if that makes sense]) I'm already able to think of a number of interesting things to do with it.

The first is to attempt to find the 'centre' of the UK population. This is a complicated thing to define as there are several contradictory ways in which you could choose to define this, for instance as simply the 'mean' location in population (i.e. as a weighted centre of gravity definition) - this yields a centre a little to the North East of Birmingham.
Other ways of defining it may be as a function of radius - for instance popdensity*radius will give a point where, subject to certain horrendous assumptions, all of the UK could drive and minimise the total time. This turns out to be a similar point to the previous. Finally suppose that you want to build an attraction or business to which having to travel a distance will put people off. Here one would get something like inverse radius relationship (distance places are less attractive) and the point of maximum popdensity/radius would give you the best location for your site. Unsurprisingly this turns out to be in London.

I'd like to do other things with the data, but it will take me a little while to get around to them - for instance:

i) What would be a mathematical best transportation network covering the country?
Assume that all journeys are person-to-person. The number of people going from and to a given point is thus simply a function of population density. We can define a function, probably of the form 'a/r^{b}', to give a 'benefit' function of the benefit of having a transportation network nearby, and a 'cost' function of something like 'c*d', where d is the length of the network. By maximising the benefit relative to the costs we can get an idea of where in the country it would be best to build our network.

ii) 'Fair' constituencies.

Consider a parliamentary constituency. To ensure that no party is offered an advantage by the placement of boundaries it would be best to define some kind of (mathematical?) definition of what makes a fair region. Ideally we would want each constituency to have the same population, for each to be as compact (i.e. have as small as possible boundary given the area) as possible (to reduce transportation costs for candidates) and for boundaries to pass through regions of as low population as possible (to prevent towns being arbitrarily split in half). We can again write a cost function, running it would give us a 'fair' division of the country.


Any other ideas?



The 'centre of gravity' of the UK population is located just to the North-East of Birmingham.


If you want a site which is close to a lot of people, the London seems to be the way to go.

( Who comments the most on this journal? )

I suspect that the results would be somewhat different without my game of 'Mafia' - speaking of which - does anyone fancy starting another game on-line, or meeting up one night (Birmingham people) to play for real ;).

I'm currently watching 'Jamie's Fowl Dinners', a description of the Chicken and Egg industry in this country. If anyone has missed it then it's bound to be repeated sometime on More4 and I highly suggest watching.

Anyway - the thing which interested me was the attitude of Supermarkets and how it accords almost entirely with my own perceptions of them, for instance in terms of farmed Chicken for meat the situation of the supermarkets was pretty much as follows:

* Waitrose is almost already 'there' with all their branded Chicken produced in properly farmed sites - although they do still use a fast growing breed.
* Sainsbury's has made a commitment to changing their practices, and hopes to bring all their Chicken up to a 'Freedom Food' standard.
* Tesco acknowledges that there are welfare problems, but is aware of the issues faced by poor families in terms of buying meat.
* Morrisons is happy to stock a range of differently farmed Chickens.
* Asda refuses to acknowledge that there is even a problem.

Maybe I'm just projecting here, but its interesting how sometimes prejudices about companies do appear to be justified.

A quote from New Scientist:

Gay Switch:

Homosexuality in fruit flies can be switched on and off by using drugs to alter the strength of nerve-cell junctions in male flies' brains. The changes makes males unable to sense a pheromone that normally stops them mating with other males (Nature Neuroscience, DOI: 10.1038/nn2019)

Is it just me or is this bad reporting. I mean, there are two issues I primarily have with it:
# It implicitly assumes that Homosexuality means having sex with the same sex: This is a somewhat outdated attitude isn't it? I mean its one of those things whereby a homosexual is a person who chooses* to have sex with people of the same sex, but someone who has sex with people of the same sex is not necessarily a homosexual. The flies here would appear to mate with both males and females which in my book would (without my second issue below) make them bisexual.
# Its not about homosexuality anyway - my reading of this article is that the chemical changes prevent the male from determining the sex of an individual. As sexuality is explicitly dependent upon sex and gender if one cannot recognise gender then one surely cannot have a sexuality? If I were to remove the ability of a fruit fly to discriminate between living and dead flies, would I be able to conclude that this led to necrophilia?

Anyway. Meh.

Neil.

* N.B Chooses here being used in a somewhat subtle manner to mean 'is not compelled by circumstance', therefore excluding rape and those who manifest homosexual behaviour in extreme environments such as prison. I don't take a stand on nature versus nurture origins of homosexuality, but acknowledge that it is not a 'choice' in the same way that becoming a vegetarian is a choice.

I continue to be amazed that people's voices almost entirely fail to match up with the image I expect of them - Kenneth Horne ('Round the Horne' et al; sixties radio shows) looks like this: http://en.wikipedia.org/wiki/Image:Kenneth_Horne.jpg

I wasn't expecting bald, was expect younger, and - I dunno, he's got such a smooth voice I was expecting a smoother presentation (if you know what I mean).