ptc24: (Default)
2015-07-26 04:18 pm

Pansonic Lumix DMC-TZ60

Time to review my other main camera. The Panasonic Lumix DMC-TZ60 is the camera that rekindled my love of photography - it's also the camera that caused me to buy a lot more cameras in 2014. Go figure. Anyway, as with my other camera, this review is as much about this sort of camera as it is about the specific model.

Note that the TZ60 is called something else in the USA - the ZS40 if I recall right. Anyway:

review )
ptc24: (Default)
2015-07-25 07:42 pm

Panasonic Lumix DMC-GF6

People like to ask me questions about my cameras - in particular, there are two which get attention in particular. In this post I'll deal with the Panasonic Lumix DMC-GF6. There are lots of reviews on the web, but I'd like to give my perspective, with a particular view to buying advice.

My background, and my thoughts on the camera and similar cameras )

And a P.S. )
ptc24: (observer)
2015-07-14 10:44 am

Pluto - maybe a planet after all?

There's an article in Vox on Pluto, and there were some graphs where I thought, "log scale, dammit!"

So I found the masses of planets, and known dwarf planets of known mass, and did a plot:

Note that the x-axis is on a linear scale - not a log scale. With a log scale you don't get a nice straight line - not really. You can fit it to a power law but the fit isn't nearly so good, and there's obvious curvature:

Thing is, lots of things fit power law distributions like this. Most famously, words. But also, asteroids. Kuiper Belt Objects are harder to observe than asteroids but the same power-law statistics might apply to them too. So there's something different between "big 8 planets + dwarf planets" as a group and asteroids as a group, or KBOs as a group.

Note that the exponential law fits better than the power law if you exclude the dwarf planets too - it's not an artefact of welding two groups of dissimilar objects together. Likewise I only have four data points so take this with an astronomical quantity of salt, but the exponential fits the dwarf planets alone better than the power law - and the power law fits better if the objects have ranks 9, 10, 11 and 12 rather than 1, 2, 3 and 4. So "big 8" planets and dwarf planets fit together nicely as a group.

Of course, by contemporary definitions of "planet", this is all irrelevant - they have to have cleared their orbits. However, maybe the current definition isn't the best definition, and maybe one that includes dwarf planets "carves nature at the joints" better.

Incidentally, the data I collected from Wikipedia:
  Rank Mass
Jupiter 1 317.8
Saturn 2 95.2
Neptune 3 17.2
Uranus 4 14.6
Earth 5 1
Venus 6 0.82
Mars 7 0.11
Mercury 8 0.06
Eris 9 0.0028
Pluto 10 0.0022
Senda 11 0.00022
Ceres 12 0.00015

ptc24: (tickybox)
2014-04-02 07:50 pm

I saw a couple of polls on the YouGov site, and wondered about having a go

Poll #15172 Polls nicked from YouGov
Open to: Registered Users, detailed results viewable to: All, participants: 13

What is your opinion of the font Comic Sans?

View Answers

Like Very Much
0 (0.0%)

Quite Like
2 (15.4%)

Don't Like
7 (53.8%)

Can't Stand
2 (15.4%)

Don't Know
2 (15.4%)

Do you think dreams do or do not reveal something significant about your life?

View Answers

4 (30.8%)

Do not
5 (38.5%)

Don't know
4 (30.8%)

ptc24: (Default)
2014-03-19 06:35 pm

Analogy analysis

So my source was CopyCat, I was recently reading a book by Melanie Mitchell (Complexity: A Guided Tour) - she was the PhD student who did the work in Douglas Hofstadter's group. (Incidentally... somehow I feel I should get on better with Hofstadter than I do. There's something about his style or something that makes me less enthusiastic about him than I'd normally be for someone of his approximate description. I don't know...) She had a chapter where she got to describe her PhD, it sounded interesting. One of the interesting things was looking at some of the human results from this sort of thing.

There's a paper here
which suffers from the odd problem of sometimes rendering "ij" as "y" for some reason, which is a pain as it comes up a lot in the examples.

There are more examples here - it's interesting looking at the sorts of answers people give, and the sorts of answers that well-developed computer systems give at the end of their run. Best to do some examples, and make my points as I go.

Here be spoilers )

Anyway, analogy. It's sort of like induction but more so. It's unreliable and variable and has all sorts of subjective properties - but you really can't do without it at all, and there are a lot of regularities there too - at least if you're willing to count strong-but-not-absolute biases as "regularities" (ETA I should just mention here that there's no induction without inductive bias - there's another place we might make make an analogy-induction analogy). One particular application of analogies is in understanding other people - I think it's easier to understand and get on with someone if their analogies work the same way yours do. Analogies in moral reasoning I think are particularly interesting, but maybe another time.

There's a small point I'd like to make about culture. On the one hand, I can see plenty of room for cultural biases. On the other hand, one needs to be very very careful here lest one starts perpetuating some rather silly views. One thing to think about is what I call the "culture acquisition problem" - how do we learn our culture - most notably our language? If we're learning it by induction and analogy, then there must be, or have been, some original part of us, that induces and makes analogies while untainted by culture.

Going whooooooooosh! far off into the distance here: I had a thought that said, "If materialism is true, we can learn something about matter just by thinking". Of course this requires very loose definitions of most of the words after the comma, and I get annoyed when other people do that...
ptc24: (tickybox)
2014-03-17 06:07 pm

A little analogy-making game

Based on some research I read recently. I'll post more about it when people have had the chance to have a go.

Poll #15102 Analogy-making game
Open to: Registered Users, detailed results viewable to: All, participants: 13

Suppose the letter-string "abc" were changed to "abd"; how would you change the letter-string "ijk" in "the same way"?

OK, now try aabc -> aabd, ijkk -> ?

abc -> abd, kji -> ?

abc -> abd, mrrjjj -> ?

abc -> abd, rssttt -> ?

abc -> abd, xyz -> ?


View Answers

I'm too familiar with this work to answer this honestly
3 (23.1%)

I remember this, but not the specifics
5 (38.5%)

I sort-of vaguely remember something like this, or was it something else?
0 (0.0%)

This is new on me
7 (53.8%)

2 (15.4%)

ptc24: (Default)
2013-10-09 11:45 pm

Nobel Prize in Chemistry 2013

There was a Nobel Prize in Chemistry "for the development of multiscale models for complex chemical systems". This is tangentially related to my PhD, so it seems like something I can have a go at explaining aspects of it.

Proteins often bind to small molecules. Some are receptors that bind to signal molecules, some are enzymes that bind to the transition states of the reactions they catalyse, some are transport proteins, some may have other reasons for binding to things. From my point of view as a sometime supramolecular chemist, there are two questions. Question 1: How come proteins are so big? Question 2: How come the binding constants are so high?

With Question 1 - if you think about all of the bits of the protein that engage in intermolecular interactions with the "guest" molecule, then that's quite small, and if you think about the smallest thing that would tie those bits into a single molecule, that's quite small too. You'd end up with something with a radius... roughly a third of the size of the radius of the average protein maybe? Now maybe proteins are just really inefficient, but that seems unlikely, especially with Question 2. It turns out that we can make artificial "host" molecules for these "guests", often using standard synthetic chemistry techniques, which are nice and space-efficient in that they contain all of the groups that bind to the guest and a small amount of other stuff to make all of those groups into a single molecule (often a big ring with a hole in the middle - a "macrocycle"). There's a problem, though; these synthetic host molecules have much lower host-guest binding constants than their protein counterparts - the binding constants are orders of magnitude lower. So maybe all of those bits of protein are doing something after all.

There was one afternoon in the Chemistry department, when I went to see a research lecture, and they were talking about "QM-MM" - a computational chemistry technique designed specifically to address the issues with proteins binding to small molecules and catalysing reactions. The problem is as follows - you want a really good model of the binding site. To get this, you need to do quantum mechanical (QM) calculations directly. Now you can't solve the equations analytically for anything more complex than a hydrogen atom, and you can only really get numerical solutions that converge exactly for things about as complex as a helium atom (maybe things have moved on a bit since my lectures), but there are some simplifications which make calculations feasible for much larger systems. However, these calculations are still computationally expensive and don't scale well - if you double the number of atoms (really, the number of electrons) in a simulation you more than double the computational demands. So simulating a whole protein like this is a non-starter.

Enter Molecular Mechanics (MM). Here, you back off to classical mechanics. You treat each atom like a ball attached to other atoms via springs; there are some simple equations to say what the relevant forces are for each atom-atom distance. You estimate the parameters for these equations "empirically" - usually this involves doing some QM simulations of small systems, and then fitting the parameters to that. Once you have your MM model set up, you can do calculations *much* faster than QM calculations, and I think they scale better too. One thing you can do with this is Molecular Dynamics (MD) - you calculate the forces on the various atoms at a given time, use that to update the velocities of the atoms, use that to calculate where the atoms will be a few nanoseconds (or whatever) later, then rinse and repeat. You can have a go at simulating protein folding using these techniques, although the last time I was listening to research lectures, it seems there was a long way to go before you could just crank out highly accurate protein structure predictions. One problem with MD is that MD simulations tended to consistently gain or lose energy due to the various approximations involved - not something you want!

So, to try to understand protein hosts binding to small molecule guests (and in some cases catalysing reactions), you'd like to do a simulation. Nothing less than QM calculations will do for the binding site. For the rest of the protein, hopefully you can get away with MM calculations. But what about the bits where the bits you're simulating with QM meet the bits you're simulating with MM. The answer seems to be "lots of complicated maths and physics", and it seems you can get a Nobel Prize for coming up with a sensible way to do this.

So, do these techniques give good predictions to do with proteins binding to small molecules? Insight? Do they help us to explain things? Alas, I don't know the answers to these questions, but we can hope. And, y'know, it's only the Nobel Prize. It's not like having an element named after you - now there's a real honour.
ptc24: (Default)
2013-09-23 07:05 pm

Bi visibility jacket

OK, that was just a pun, I haven't made one or even a picture of one, you'll just have to imagine it.

Anyway, I suddenly feel inspired. From Chapter 1 of Getting Bi, this came to mind:

I used to define bisexuality as "the potential to be attracted to people regardless of their gender." Then one day I was chatting with my friend Alberto, who, like me, identifies as bisexual. I blithely stated my definition and he looked at me incredulously.
"Regardless of gender? No, no, no! There's no 'regardless' about it for me. For me it's all about difference. I'm attracted to cheerleaders and football players. It's precisely the extremes of difference that attract me." He looked me up and down, smirked, and said, "Robyn, you would be way too butch for me!"
I threw a pillow at him, he threw it back and we laughed. But I learned from that conversation. Some of us who identify as bisexual are in fact "gender-blind." For others—in fact for me—it's androgyny or the blending of genders that compels. Still others, like Alberto, are attracted to the poles.
Then, to complicate things further, I learned a lot from my intersex, genderqueer and transgender friends.

Getting Bi is a fine example of a fine sort of book - an anthology of short, mainly autobiographical, pieces. There's a lot of diversity in there - in terms of how people's sexuality works, how their identity works (there as a few "I don't identify as bi as such but I'm close enough that I'm happy to contribute to this anthology"), how they define bisexuality, what else is going on in their life and how it all interacts with their sexuality and identity and all the rest.

Anyway, I had the idea of a bi visibility jacket, how about a bisexual umbrella to go with it?
ptc24: (Default)
2012-08-30 10:22 am

Does anyone here know Latin

So I wanted some Latin for "out of thin air" - basically I wanted a more scathing version of ex nihilo or a priori. Google translate suggested ex tenues auras, and I like the implication of "tenuous", but I'm not sure it's a) even remotely correct and b) whether there's an actual Latin idiom that carries the same air of dismissal, and which doesn't make you look stupid for using it (and lead to people pointedly wondering where you got that phrase from).
ptc24: (Default)
2012-06-06 08:57 am

Go links

We were talking about Go recently, and I realised I had a file sitting somewhere with some useful links in it - in particular I remembered igowin being good and useful for getting off the ground. So here it is again, back from 2005:

Starting out with Go )
ptc24: (tickybox)
2012-04-20 10:19 am

The meaning of "knowledge"


In Town X, a burglary is reported to the police. They investigate, and their investigation leads to Bob, who has previous convictions for that sort of thing. They find the stolen goods at Bob's house, Bob is questioned, he confesses, it goes to trial, Bob pleads guilty, the jury is convinced by the evidence, and Bob is sentenced accordingly.

In fact, Bob had not committed the burglary. In fact, he had a provable alibi - he was out committing another burglary (of equivalent severity) elsewhere in town at the time. The original burglary was done by a friend of his, and he was just storing the goods at his house. Given that demonstrating his alibi wouldn't have saved him any prison time, and would have involved fingering his accomplices (which would have damaged his standing in the criminal underworld), he felt that the simplest thing to do was to take the rap for the original burglary.

This is all to illustrate a philosophical point about the word "know" (and "knowledge" etc). A poll:

(ETA Note that the previous convictions don't count when I'm asking whether the legal system knew Bob had committed burglary... I meant to ask "...committed burglary on that day" or something similar. This ETA is before anyone but me has answered this, so all the results reflect this ETA)

poll under cut )
ptc24: (Default)
2010-12-05 03:44 pm

Chomsky was wrong!

There's a classic example sentence he has, "Colourless[1] green ideas sleep furiously", which he holds up as an example of grammatical gibberish, and the contrasting "furiously sleep ideas green colourless", which is an example of ungrammatical gibberish. The difference between nonsense and word salad, if you like.

However, I've realised that you can add punctuation, and get "Furiously sleep ideas green, Colourless!", which is grammatical, and you could even imaging a hypothetical scenario for it. That scenario would probably be an MMO, a bit like World of Warcraft. The interpretation goes as follows:

"sleep" has a transitive sense, meaning "to cast a sleep spell on". Once you know this it's obvious that the sentence is an imperative sentence, and everything else falls into place quite quickly. To sleep ideas green is presumably to repeatedly cast the sleep spell on the ideas (presumably mobs called "ideas" - stranger things exist in WoW) until they turn green - after all, you can shout yourself hoarse. Presumably this is best done in a furious manner. And, well, people can have all sorts of bonkers handles in MMOs, so Colourless is not nearly as crazy as some I've seen.

In fact this all feels less strained than coming up with a good scenario for the allegedly more grammatical reverse statement.

[1] This was probably actually "colorless", but whatever...
ptc24: (Default)
2010-11-06 08:13 pm

Books (well, book)

Newton and the Counterfeiter, Thomas Levenson

Some nice light historical reading, the nonfictional counterpart to Neal Stevenson's Baroque Cycle, but one sensible-sized paperback and not three bricks. The counterfeiter is William Chaloner, who seemed to have amazing amounts of chutzpah and persistence, and kept coming up with all sorts of crazy criminal schemes.

One interesting thing seems to be that hunting counterfeiters was something that Newton needed to be pushed into. His post as Warden of the Mint was a sinecure, and it was a bit of a surprise to all concerned when he actually applied himself to the problem of getting the mint to run at the capacity needed to recoin all the silver in circulation. However, during this someone pointed out that the Warden was meant to be in charge of investigating and prosecuting counterfeiters - this was something Newton didn't want to do, but his arm was twisted, and he turned out to be good at it.
ptc24: (Default)
2010-10-31 01:45 pm

Headphone cable

I've got a nice set of expensive headphones, with a problem - the left headphone has gone very quiet. They're not cheap to replace, but they do have a detachable cable, so I suspect the problem might be with that. Before I go and buy one, does anyone have a headphone cable I could borrow for diagnostic purposes?
ptc24: (observer)
2010-10-19 02:17 pm

(no subject)

Units quiz

It's a BBC science quiz - I got 9/9, although some of those were intelligent (or lucky) guesses or deduction-by-elimination rather than by directly knowing the right answer.
ptc24: (Default)
2010-10-17 08:29 pm

Dimensions in popular culture

No, not alternate realities, actual dimensions of things. A 2x4 is useful for hitting people with, and of course everyone should know the significance of 1x4x9. But surely there are others that I can't think of...

ETA: OK, there are lots of things which are kind of one dimensional. But how about things which are 2D or 3D like the examples above (or tim's example)?
ptc24: (tickybox)
2010-10-07 03:52 pm

(no subject)

Poll #4686 Equality and Hierarchy
Open to: Registered Users, detailed results viewable to: All, participants: 19

A has authority over B. Can C interact with both A and B as equals?

View Answers

7 (36.8%)

0 (0.0%)

sort of
1 (5.3%)

5 (26.3%)

under some circumstances
13 (68.4%)

other (please comment)
0 (0.0%)

I have some nifty maths about (in)equality which I will explain in a comment
0 (0.0%)

lowercase tickybox
0 (0.0%)

ptc24: (observer)
2010-09-28 09:47 am

Imagine the scene...

...adrift at sea, nothing but ocean all the way to the horizon in every direction, no supplies, no hope of rescue, not even a proper lifeboat, just some driftwood lashed together, and only a distinctly taciturn Protoss for company.

Ladies and Gentlemen, I bring you Stark Raft.

(or maybe not, actually writing the story would be too much like effort)