Wikipedia:Reference desk/Archives/Science/2016 November 30

Source: Wikipedia, the free encyclopedia.
<
Science
Science desk
< November 29 << Oct | November | Dec >> December 1 >
Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.


November 30

Set in the stars....

Whether your favorite iconic image is Trump, Castro, or Obama, if it can be drawn with simple, heavy lines, it seems like in theory, a field of stars might line up with that artwork in a suggestive way. As a very large number of stars have been mapped by now, I'm thinking that for each of these folks there may be a constellation waiting to be promulgated, even if it might be a very small one. But has anyone designed a publicly available interface by which a piece of submitted art can be searched among all the stars at all possible scales until something abnormally suggestive is uncovered? Is there a statistical analysis possible for how good or bad a match that might turn out to be on average, given available data? Wnt (talk) 02:43, 30 November 2016 (UTC)[reply]

Just for reference, this is a rendering of every star visible to the naked eye, in a light-pollution free area, from...somewhere, not really sure. Now, this is a bit less than you'd be concerned with, since this is everything you'd be able to see standing still, not counting what stars might be behind you, etc. Someguy1221 (talk) 02:50, 30 November 2016 (UTC)[reply]
Sorry, to be clear, I didn't just mean a search for small constellations of the desired shape, but also faint ones - even if a constellation is only visible through a powerful telescope, there is still substantial amusement value to be had if someone found it and posted an image, where people could say "oh yeah, that looks like him" or have fun peering at adjacent stars and saying "you know, I think there's somebody over there behind Kennedy on that starry knoll..." Wnt (talk) 12:08, 30 November 2016 (UTC)[reply]
A bit of terminological pedantry: what you are describing would be asterisms, not constellations. {The poster formerly known as 87.81.230.195) 176.248.159.54 (talk) 14:14, 30 November 2016 (UTC)[reply]
You may be interested in
the Coathanger. Sagittarian Milky Way (talk) 15:00, 30 November 2016 (UTC)[reply
]
If you mean that anyone looking at that set of stars would say "Wow, that's person X !", I doubt that is the case. Even most of the current constellations really don't objectively look like what they are said to represent, with a few possible exceptions, like the Big Dipper, and that's only 7 stars. To make a convincing representation of a PARTICULAR person, you would need far more stars than that, and without extraneous stars visible in the same field. On the other hand, for a generic "smiley face", perhaps as few as 5 stars could work, with 2 for eyes and 3 to form the mouth. StuRat (talk) 15:17, 30 November 2016 (UTC)[reply]
Well, there are something like a billion catalogued stars at present, vs. 7000 naked eye stars. This means that I'd expect the best possible match, if obtainable, to be somewhere like 130,000 "better" than the best constellation you could find looking about by eye. Now how exactly "better" translates in terms of ... some measure ... that I can't say. Wnt (talk) 17:47, 30 November 2016 (UTC)[reply]

What's the fewest irregularly placed dots needed to form a face recognizable as a particular person ?

To go along with the above Q, does anyway have examples of this ? I suspect that some faces would need fewer than others. For example, Groucho Marx might need very few, just enough to form dark eyebrows and a cigar. Somebody with a less distinctive face may need far more, to distinguish them from other similar people. StuRat (talk) 17:42, 30 November 2016 (UTC)[reply]

[1] Blooteuth (talk) 18:32, 30 November 2016 (UTC)[reply]
Not sure who that is, Charlie Chaplin ? StuRat (talk) 03:15, 2 December 2016 (UTC)[reply]
The sample image made of 2300 white dots is recognized as the iconic face of Charlie Chaplin but the technique is inadequate to distinguish faces generally. A similar number of dots could be placed with more realistic effect using a dithering algorithm such as Floyd–Steinberg. The example is limited to the dot arrangements on domino tiles. These examples of Pointillism are best viewed from a sufficient distance to lose visibility of individual dots. One can then describe the resolution of the Chaplin example as a 20 x 23 array of pixels, each of which has an average brightness of one of 10 levels. As a binary file (with no attempt at data reduction) the image could be stored as 25x25x24=16384 bits. In practice, facial images in monochrome are seldom shown as binary images but more often with much better fidelity either by halftoning in print (uses a fixed array of dots whose sizes are continuously variable) or by pixellated displays with a given Grayscale. Blooteuth (talk) 16:51, 2 December 2016 (UTC)[reply]
Good info, thanks. Now for the hard part. What's the chances of finding 2300 stars so arranged, with no other stars in the field of view, given even the most powerful telescope we have ? I'm thinking essentially zero. StuRat (talk) 17:01, 2 December 2016 (UTC)[reply]
The Scream
Here's a class report, in which one student set out to answer a variation of your question (using a bit more technical terminology): "How many eigenvectors (eigenfaces) are required to recover an original face by some percent of the total?"
He presents an overview of the math involved, reviewing some of his cited literature: Face Recognition using Principle Component Analysis
  • Turk and Pentland's 1991 paper from MIT is often considered the "grandfather" of this line of research (PCA for face classification algorithms).
    Principle component analysis
    is the branch of applied mathematics techniques that let us find the smallest number of elements (in some abstract coordinate space) that fully and uniquely describe a data set; it's essential to modern technologies like video compression, video analysis, machine vision, and so on.
Now, you're using spare dot patterns (rather than eigenfaces), but the math is the same. Are your dot patterns
orthonormal and spanning? If not, it's going to be very hard to formally answer your question without loss of generality
.
Nimur (talk) 19:31, 30 November 2016 (UTC)[reply]

Borked Latin taxonomic names

It seems that ungrammatical Latin taxonomic names (like Baracktrema obamai instead of correct Latin genitive obamae) have been around for at least several years. Presumably scientists are supposed to know basic Latin to give correct names. Is it centuries-old or a relatively recent phenomenon? Brandmeistertalk 11:50, 30 November 2016 (UTC)[reply]

It's been a complaint since at least the 1890s. So, depends on your definition of "recent". --Jayron32 11:56, 30 November 2016 (UTC)[reply]
Thanks, didn't know it goes back to the 19th century. Weird. Brandmeistertalk 12:01, 30 November 2016 (UTC)[reply]
Note also that the name "Obama" is not Latin: it is from a Kenyan dialect,
second declension, Latinizing by adding -us to the name to make the nominative Obamaus, then the actual form obamai is correct. And this becomes more a Language question than a Science question. --76.71.5.45 (talk) 13:19, 30 November 2016 (UTC)[reply
]
Only tangentially related, but also check out This common mistake which also involves bad latin. --Jayron32 13:42, 30 November 2016 (UTC)[reply]
Recall that, generally speaking, binomial nomenclature doesn't really have anything to do with Latin per se, people can and do put in any kind of nonsense they want. Binomial_nomenclature#Derivation_of_binomial_names lists some non-latin, non-greek examples. Also, while some taxonomists may be a bit sharper on their classics, modern scientists in the USA generally receive no mandatory formal training in Latin or Greek, though some of us did choose to pursue it. SemanticMantis (talk) 16:05, 30 November 2016 (UTC)[reply]
See Wikipedia talk:WikiProject Plants/Archive65#Additional information about Hydnocarpus wightiana (December 2014).
Wavelength (talk) 18:38, 30 November 2016 (UTC)[reply]
It is a constant source of amusement to speakers of a language how it is mangled by foreigners - for example São Paulo rendered as São Paolo. One area of debate on how names should be rendered into Latin concerns the spelling of the names of rulers. Following a line of authorities from JACOBUS to GEORGIVS some people wanted the present Queen's name on coins to be spelt ELIZABETHA. However, it was pointed out that the Hebrew nominative is unaltered in Latin (and it could have been mentioned that the coins of Elizabeth Tudor showed no variation in the name), so ELIZABETH it was. 81.134.89.140 (talk) 01:24, 1 December 2016 (UTC)[reply]

Elements 119-120

Now that all of the first 118 elements have been named, are scientists working some on trying to create atoms of elements 119-120?? If not, please explain what they're waiting for. Georgia guy (talk) 14:40, 30 November 2016 (UTC)[reply]

They are! But new technology is needed and 119 and 120 are stretching the limits of current technology. We should be able to get to these elements in the next five years (though it could be as soon as next year or as late as in ten years). Double sharp (talk) 15:00, 30 November 2016 (UTC)[reply]
See the Wikipedia articles titled Ununennium and Unbinilium. --Jayron32 17:25, 30 November 2016 (UTC)[reply]
It might be more interesting to see different isotopes of previously identified elements. It can be frustrating because some tables - including in Wikipedia - don't clearly distinguish between isotopes that are not stable vs. never plausibly attempted. But as I recall, there are several of the "known" heavy elements where the heaviest isotope that has so far been created is also the most stable.
Oh, they're working on that too. Already the synthesis of 284Fl was a preliminary attempt to fill in the big gap between isotopes made by cold fusion (neutron-poor, like 278Nh) and by hot fusion (neutron-rich, like 286Nh), except that it wasn't nearly as impressive since it instantly underwent spontaneous fission. But yes, there are plans for these. If we could get the 2n evaporation channel for 249Bk + 48Ca or 248Cm + 48Ca, as well as use heavier Cf isotopes like 250Cf or 251Cf, we could immediately get 293Og and 289Lv on the neutron-poor side, as well as 295,296Og, 295Ts, 294Lv, 291Mc, 290Fl, 287Nh, 286Cn, 283Rg, and 282Ds on the neutron-rich side. (Some of these, like 291Cn, may have significant EC branches leading straight to the predicted centre of the island around 291,293Cn.) Further work using light isotopes with 48Ca projectiles, such as 241Am and 239,240Pu, should give 285,286Mc, 283Fl, 281Nh, 279,280Cn, 277Rg, and 275,276Ds. Of course, the problem with these is that the fission barriers gradually disappear: soon there is no barrier to shift the nucleus to a non-spherical shape so that it can undergo SF. Also the cross-sections of the reactions with 48Ca decrease exponentially as the neutron number of the target decreases. But all of these are reachable with current technologies, unlike 119 and 120, for which the cross-sections with 50Ti projectiles are going to be too low and will require upgrades. (Unless we, against all odds, managed to make a target of 254Es; I'm not even thinking about the possibility of 257Fm, since it is too outlandish.) Double sharp (talk) 16:15, 4 December 2016 (UTC)[reply]
I should also add that due to the importance of these isotopes in the initiators of nuclear bombs (sort of a "blasting cap" if you will) I suspect we're not being told every last little thing known. Wnt (talk) 17:52, 30 November 2016 (UTC)[reply]
Also, George, I'd like to respond a bit to the "explain what they're waiting for" bit, as it sounds like you may not understand the process under which science is done and scientists work. There aren't really avenues that scientists can explore (ethically, so I'm discounting things like just cutting off limbs to see what happens or challenges of human cloning) that they are just "waiting for," and not exploring. If they aren't doing something, there's probably a good reason. If it is beyond the limits of current technology or understanding, you can bet they aren't just "waiting," but rather are working to develop the needed technology or trying to discover the knowledge. That said, they need funding and support to do so. The lab I work in would certainly turn out results a lot faster if we could afford a new spectrometer, for example, but we don't have half a million dollars lying around to just spend on that. --OuroborosCobra (talk) 19:42, 30 November 2016 (UTC)[reply]
Shorter: They're waiting back to hear whether their grant resubmission received a high enough score (that is, assuming it is within the "outstanding" category; if it is merely "excellent" they need not wonder) to get provisional funding, provided the budget isn't cut (which it will be). Wnt (talk) 20:12, 30 November 2016 (UTC)[reply]

Genetics or disease?

What is the cause of extremely large buttocks on women if they dont over eat? Im talking at least 100" around. I saw one woman with this condition and wondered what causes it/ THe rest of her body seemed fairly normaL size but the buttocks wre HUGE?--178.106.234.63 (talk) 20:24, 30 November 2016 (UTC)[reply]

see Steatopygia. Dr Dima (talk) 20:32, 30 November 2016 (UTC)[reply]
Female sex hormone causes fat to be stored in the buttocks, thighs, and hips in women." SemanticMantis (talk) 20:48, 30 November 2016 (UTC)[reply
]
It is worth following the
Khoikhoi people, a subset of Khoisan people. This is of course not the only race seen to have bizarre features by another, which is an interesting development both in terms of sexual selection[2] and perhaps in view of a potential role of assortative mating in beginning a movement toward future speciation; in particular, when the populations are exposed to each other, sympatric speciation. (I am not arguing by this that any such speciation event is yet underway, or ever will happen; only that when we look at these differences perhaps we get an intuitive insight on what speciation feels like for the populations involved. Is steatopygia repulsive or alluring? As you look at women like this more and more, how does your response to them develop? Does culture mean some people will never accept the trait while others will favor it, and if so, does that mean that culture can determine whether speciation occurs? Etc. You can think up questions to ask as fast as you can type, but it is deucedly hard to answer any of them.). Wnt (talk) 20:54, 30 November 2016 (UTC)[reply
]
The answer to your final question is "yes", provided that cultural norms can result in reproductive isolation. That claim is indeed hard to prove, but we do have some decent evidence. See the final two refs at sympatric speciation for what I think is some of the best contemporary evidence for sexual selection driving speciation. SemanticMantis (talk) 21:01, 30 November 2016 (UTC)[reply]
Oh my, that was an interesting find! And not what I'd expected - from a few prior examples in insects I was expecting a large inversion to rule the day, not multiple small regions. Wnt (talk) 23:53, 30 November 2016 (UTC)[reply]

(OP) She was similar to Baartmaan but I think a bit bigger and rounder. Her mother was of fairly normal size. — Preceding unsigned comment added by 178.106.234.63 (talk) 22:09, 30 November 2016 (UTC)[reply]

I am not sure if the OP is simply a racist troll (see the latest season of

Souf Pahk) or historically ignorant. See Sir Mix a Lot. μηδείς (talk) 05:14, 1 December 2016 (UTC)[reply
]

How can my posts be considered racist? I didnt mention the race of the woman I saw. She was white European actually so her race has no bearing on the reason for this condition. 213.205.252.104 (talk) 18:03, 1 December 2016 (UTC)[reply]
Make that sexist if you like, but asking what is wrong "disease or genes" with such people is a very parochial view to say the least. μηδείς (talk) 22:46, 1 December 2016 (UTC)[reply]
At no point did the OP suggest that the person or condition was "wrong", nor did they mention "diseases or genes", they merely asked the possible causes of something unusual. Are we not allowed to have any interest in the multifariousness of humanity? Putting words into another's mouth and then criticising them for those words seems to me to be sailing close to trolling in itself. {The poster formerly known as 87.81.230.195} 176.248.159.54 (talk) 23:07, 1 December 2016 (UTC)[reply]
To be clear, I tried hatting this, as casting such aspersions has no relevance to the business of answering questions but Medeis removed it. [3] And I was first to cite Sir Mixalot. I see no benefit in getting into editorializing - some people are going to love this trait and some are going to hate it, butt it is what it is. Wnt (talk) 01:21, 2 December 2016 (UTC)[reply]
OP here. Ok, lets generalise removing any contentious assumptions. What is the cause of extremely large butttocks and fat legs of people of either gender, any creed, religious belief, lifestyle, and of any race? Is it genetically inherited, or is it some sort of disease?213.205.252.104 (talk) 01:39, 2 December 2016 (UTC)[reply]
Your question was answered in the very first response you got. You needn't ask it a second time. --Jayron32 02:07, 2 December 2016 (UTC)[reply]

Light bulb life

I was about to buy these light blubs[4] until I read the fine print: "Rated Life: 750 hours". Is that a typo or what? That's a little over a month. ECS LIVA Z (talk) 22:05, 30 November 2016 (UTC)[reply]

Looks correct. That bulb at other sites also lists 750 hours, as does the 150-watt version of the bulb. Incandescent bulbs generally seem to be listed as around 1000 to 1200 hours average life, depending on which search result you prefer (here's energy.gov); it seems reasonable that a relatively high-wattage bulb would fall toward the lower end of the range. — Lomn 22:13, 30 November 2016 (UTC)[reply]
30% of the reviews are 1 star out of 5. I would probably look for a different bulb. Vespine (talk) 23:01, 30 November 2016 (UTC)[reply]
But did you read the bad reviews? It actually looks excellent. The reviews are neatly bifurcated into excellent and terrible, and specifically the bad reviews all state the bulbs were DOA. That's perfect. Either your bulbs will be broken before you open the box, or they'll all work just fine. So either they die within the return period, or you'll be happy with your purchase. Bad reviews for electronics are probably 5-10 fold overenriched, so probably most of the bulbs will be fine. In fact, the reviews go further and seem to agree that either all the bulbs in the box work, or most fail. This is the sort of review set that I look for when comparing products. I'd sooner take a fair chance of complete, immediate, and obvious failure, over a small chance of distant and subtle failure. Someguy1221 (talk) 08:29, 1 December 2016 (UTC)[reply]
I think nobody reviews a light bulb after it burns out - the reviews won't give you an idea one way or the other.
Even taking ALL the above into account, a full 30% of the reviews being 1 start IS high, compare it to the reviews of other bulbs available on amazon.
I don't have a source to cite, but I believe that within conventional incandescent bulbs, if the life is shorter it's because the bulb burns hotter and therefore gives more light and a better (less orange) color of light. I've observed this myself when replacing the brighter light bulbs in my house. I'll probably replace them with LEDs soon. --76.71.5.45 (talk) 23:16, 30 November 2016 (UTC)[reply]


Was employed for a time as an industrial buyer. i.e., Sometimes bought light bulbs for use in industry. For example, bulbs that UK miners used on their helmet lamps had a tiny bit of krypton or xenon gas added. It made them burn brighter and last longer. They only cost a little bit more, so in the long run they where less expensive. Also did a lot of photography with high luminance photoflood bulbs which not only cost more but burnt out sooner. So, I was very conscious of bulb life – or rather the lack of. The long life bright industrial bulbs were not available retail to the general public (but you bet I took advantage for my torches) because there was less profit to be had by the manufacturers from selling longer lasting bulbs to uninformed hoi polloi. It all goes back to the Phoebus cartel to reduce bulb life. Now, manufacturers are promoting 50,000 hour LED (which equates to a normal night-time usage to over 30 years) but obsolescence is already being built in. Are your LED light bulbs burning out too soon?. --Aspro (talk) 01:07, 1 December 2016 (UTC)[reply]
All light bulbs have planned obsolescence. The could built bulbs that last essentially forever (and have, see Centennial Light), but they don't because there's no money in it. --Jayron32 02:16, 1 December 2016 (UTC)[reply]
Here is an interesting article on this very topic http://spectrum.ieee.org/geek-life/history/the-great-lightbulb-conspiracy --TrogWoolley (talk) 12:07, 1 December 2016 (UTC)[reply]
And because of the light being too orange. You can only heat a filament so hot and have it last a long time. --76.71.5.45 (talk) 03:54, 1 December 2016 (UTC)[reply]
You can also buy an incandescent light bulb with a very bright filament and a very short lifespan. They call them flashbulbs... :) --Guy Macon (talk) 05:08, 1 December 2016 (UTC)[reply]
  • The
    Centennial light has been alive for so long because (1) it is orange, i.e. less hot than one would like it (see Black-body radiation or Incandescence), but also because (2) it is always on. What kills lightbulbs is the thermal stresses due to going from room temperature to a bit less than tungsten's melting point then back when switched, which happens multiple times per day in normal operation. If there is a cartel, that one bulb is not proof of it. TigraanClick here to contact me 12:47, 1 December 2016 (UTC)[reply
    ]
You can find detailed calculations online, but many are not really reliable sources. However, there is this energy.gov page which says a LED saves $3.80 of electricity per year (give or take). If you assume you will use them for five years (again, assuming lifetime does not matter: a LED should last longer than that, an incandescent light bulb will often break sooner), it means you can pay $19 more for the LED than for the light bulb. The first LED I found for sale online is not even at that price. (OK, do the discounted cash flow calculations if you want, but it will be hard to come with a credible value of the rate of interest that warrants buying incandescent) TigraanClick here to contact me 13:22, 1 December 2016 (UTC)[reply]
One consideration to the contrary is that, here in the UK, most LED bulbs on sale are designed for different fittings than the "wide bayonet" of most traditional domestic light appliances and sockets, so in addition to the bulbs one would also have to replace each appliance or socket. This might well be economical during a general house rewiring or redecoration, but otherwise onerous.
A further factor is that, in the UK, we have in the last few years already gone through a government-sponsored push to replace incandescent bulbs with fluorescent bulbs, many of which initially proved to be too large to fit into existing appliances, lampshades etc., despite the actual wide bayonet fitting being the same; many of these early fluorescent bulbs also failed in short order, and many elderly people in particular found their colour too white and "harsh" even while their output was too dim. This fluorescent push was followed up by a more commercial promotion of halogen bulbs, which again have been difficult to find with sufficient output for many situations. LEDs therefore represent a third wave of change, following two unpopular ones, so for many the well has been poisoned. {The poster formerly known as 87.81.230.195} 176.248.159.54 (talk) 23:50, 1 December 2016 (UTC)[reply]
Yes, I've had at least a dozen CFL bulbs stop working before their claimed lifetime, possibly because they are used in a cold environment where they take a long time to reach full brightness. I haven't tried halogen, and I find that LED bulbs emit more glare than illumination, though I see that the latest more expensive ones now claim a less harsh mix of wavelengths through the use of phosphor coatings. Dbfirs 00:01, 2 December 2016 (UTC)[reply]
Do you mean on the LED or something else like the lens? Pretty much every white LED uses phosphor coatings. Nominally it might be possible to make them in other ways but almost no one ever does that. (The one exception is those RGB LEDs.) Raw LEDs have had many colour options for a very long time. I don't think it's been that different for bulbs although it's perhaps taken a while to get the right balance of size and heatsinking so the bulb doesn't die to fast. Actually from my experience it's much more difficult to get a real white (i.e. 5500K+) bulb than it is to get the yucky incandescent-emulating yellow light bulbs (3000-5000K) especially in good CRIs or from reputable companies and in MR16. In addition, it's true that getting a good light output profile requires a competent designer who choose a suitable LED, reflector and lens, and many of the earlier designs may not have been up to scratch. I would be incredibly surprised if it's difficult to get a wide bayonet fitting LED bulb, since it's trivial to get them here in NZ and we aren't exactly known for how quickly we get new technologies. It's true that it's better if you get a purposely designed LED fitting, but by now LED replacement bulbs seem to have reached a level where there's generally no reason not to use one in nearly all situations. The fact that other forms of LEDs are more common seems to be a moot point if it's trivial to get what you want. Nil Einne (talk) 08:19, 2 December 2016 (UTC)[reply]
It sounds as if New Zealand is ahead of the UK in this respect, or perhaps I have just not kept up with the last couple of years of improvements in the technology. I've seen standard BC LED bulbs on sale here recently, so I'll look at their light output and see if it's more pleasant than the earlier bare LEDs that I find harsh. I don't know anyone here in the UK who has moved over to LED bulbs yet, though there must be some. Cheap LED floodlights are common here, but their output is a high-temperature narrow-band wavelength, so the light is harsh and looks much brighter than it really is in terms of irradiance. Perhaps it's just my eyes that don't like single-frequency light? Dbfirs 09:49, 2 December 2016 (UTC)[reply]
I recall a trick from obnoxious youth... a quick and measured flick of two fingers can put out an outdoor light bulb for good, or more interestingly, massively improve its brightness for a day or two before it dies. Wnt (talk) 19:08, 1 December 2016 (UTC)[reply]
That only worked with certain designs of filament support. After the light bulb had burnt out, careful manoeuvring of the filament could make it rejoin and weld together, burning more brightly for a while before it burnt out again. Dbfirs 19:24, 1 December 2016 (UTC)[reply]
Lightbulb conspiracy makes me laugh AS IF some Chinese pirate company wouldn't have made it by now if it was so 'simple' to make a perpetual lightbulb. There's whole companies dedicated to making iPad knockoffs, but not a single factory has thought to make a lightbulb that doesn't burn out? The first company to do it would make an absolute killing. Vespine (talk) 22:16, 1 December 2016 (UTC)[reply]
Agreed. In the 1970s, a company called Wynne Industries (if my memory serves me correctly after such a long time, but not to be confused with the more recent American company that uses that name) marketed "everlasting lightbulbs" in the UK. They were "guaranteed for life". They just had a higher-resistance filament that would reach full brightness at perhaps 270 volts, so emitted less light at the UK voltage of up to 250 volts. They lasted for years, with a much yellower light than normal bulbs, but eventually did burn out, by which time the company had gone out of business. They probably did make a killing, but not in an honest way. Dbfirs 23:34, 1 December 2016 (UTC)[reply]