Added: 1 month ago
From: numberphile
Views: 22,154
Sort by time | Sort by thread (beta)

Link to this comment:

Share to:

All Comments (299)

Sign In or Sign Up now to post a comment!
  • the thing my friends showed me which was awesome is that if you take any number and divide it by 9 for the number of digits, you get the number.

    For example:

    59 / 99 = 0.5959595959595959

    143 / 999 = 0.143143143143

    123456789 / 999999999 = 0.123456789123456789123456789

  • NEIN! NEIN! NEIN! NEIN!

  • lol i thought i was on a hitler video for a second..."NINE NINE NINE!!"

  • Seems to me that we can write this pattern as a rule, sqrt[(4-2φ)*10^(2n-1)] = sqrt(6-2φ)*10^n, where φ is the golden ratio.

  • if u take the square root of 100 u get 10.0, grafting number?

  • if you notice, the second thing he tries, the number before the decimal is the original number plus one

  • woh

  • Excel??

    At least learn Python or something. You're a mathematics professor! Get some real tools.

    Not that I'm all that clever, but...

  • @TheGzeus THIS IS WHAT I GET FOR COMMENTING EARLY.

    I'M AN ARSEHOLE.

  • @TheGzeus lol

  • @TheGzeus Its the internet, seems to bring out that ass in all of us >.<

  • It seems to happen near orders of magnitude.

  • @PykohYT yeah. as long as its close enough to 10, 100, 1000 it works. sqrt100=10,0 so that's why

  • Ernest Rutherford, "All science is either physics or stamp collecting.

  • root of 100 is 10.0 OMG !!

  • lim n→infinity

    sqrt(10^(2n-1)(3-sqrt(5)) = (10^n(5-sqrt(5))

  • Comment removed

  • Didn't happen to come across the powers of ten, did you? They seem to fit the pattern described in the beginning, albeit not the mathematical formula with 3 - sqrt(5)

  • this is stupid -_-

  • @IcyCylena trollolololol?

  • @IcyCylena Actually, it seems pretty smart...

  • I wrote all this down and showed it to my Algebra teacher, and he thought I was a genius for a while.

  • @SanguineSteel What do you think the result will be if I write the 98, 99, 9998, 9999 one on the board while my Algebra teacher is out of the room?

  • You know so much random shit, lol. You should go on a game show, your a fucking walking tivia book!

  • Ugh... every time i try to go to sleep i end up watching numberphile all night instead.

  • ಠ_ಠ wtf is your problem?

  • something that could be related, but even if it isn't is interesting. with a restricted keyboard, i'll define "{}" as taking the square root of a number, do (x+{x^2+1})^n, where x is an integer greater than 4, and n is any number, you will get n repeating digits after the decimal point. so for example, (5+{26})^4=10401.99990...

    as x gets larger, the number of repeating digits increases, and the same when x gets smaller, but x=5 is when the number of digits equals n

  • 9 is a weird number...

  • Woot, he uses Python to code. Python for the win.

  • The first couple of numbers (99,98 etc.) are not even special imo, there are close to 100,1000,10000 etc. and their square roots are therefore close to 10,100,1000 etc..

  • Astonishing.

    normally I hate maths and

    English isnt my motherlanguage but I learn more then in my language

  • 1:05 sounds like inglorious bastards

  • 7:50 BLEW MY MIND. HOLY SHIT

  • I just spent an hour working on this... but it's time well spent!

    What actually happens is you can write those constants as roots of a quadratic equation, or rather a family of quadratic equations, and if you allow the number to appear deeper in the square root, there will be more constants.

    However, some of these additional roots don't give rise to grafting numbers easily! For instance, getting a grafting number from 4970-100*sqrt(2470) will be quite difficult without some programming.

  • manycam.com

  • that's cool if you want to rec your desk top or loptop you can download many cam at

  • How the fuck can you remember all those sequences!?

  • @kiemul136

    Black magic

  • Generally digit patterns don't interest me very much because they are base dependent. There's nothing intrinsically interesting about the number 98 viz a viz its square root if we are talking about digits. 98 in base 5 is still the same value, but the property being explored is no longer applicable. To find something intrinsic about numbers ignore digit patterns, find something which is true regardless of base which is just the system by which the number is being represented.

  • I programmed a little in python when I was in the 9th grade. Granted, it was a first level programming class that got cut from our school after that year so I didn't learn anything too detailed.

  • Wait, what o.O

  • Those reoccurrences of the original numbers are not freak occurrences. Ther are FREQuency occurrences.There is a differencve betwen the sqr rt of 2 and the halvsaa and doubles. When the digit is smaller than 5 the result will be a single digit. Thsi causes a repetitive sequence. This is the same law, which governs exponentation of numbers which are not 10, and having the algorithm, of number,place, and quantity match to break down. (^21 is a 21 digity number. 9^22 power is not a 22.

  • RE: Your BASE 10 reference. Which BASE number will allow PI to be a repeating number?

  • Comment removed

  • The number with the lowest accuracy in the family discussed (or rather, the smallest number) is 8. sqrt(8) = 2.82842712.

    1 is also a "grafting number".

  • he reminds me on marshall from how i met your mother :p

  • yay python!

  • i laugh at physicists who call biologists stamp collectors...i used to live with 3 post-grad physicists, and when id go to their office in the physics department there were whiteboards with various problems that id solve and get the ire of other physicists in the department, because id always sign off as the stamp collector or the lowly biologist. stamp collecting is fun, and being a plant systematist all i do is 'stamp collecting'...easy job, and i get to look at flowers all day, win-win

  • Computer programming is like skydiving in that when most people do it for the first time, they spend the entire time screaming.

    This video that shows 'recreational mathematics' is phenomenal BTW, I love recreational mathematics stuff. I wish Martin Gardner was still alive so you could do a series of videos with him...

  • up until 3:30 i was writing a comment to just make a small program that searches for them for you

  • Cube Root of 999 is 9.99...., Cube Root of 999999 is 99.9.... etc

  • i find it fascinating

  • this seems more like numerology than mathematics

  • Well, I don't know about higher maths or anything but, if there is an infinite amount of numbers, I think the probability for a random number's square root to contain the number itself is actually pretty high. Of course, this is not me taking away any credit but why is that so special?

  • is there a proof for this?

    plz reply

  • @lolz2018

    The special number he gets in the end is 3-sqrt(5)

    It can also be written as 2*(1-1/phi) with phi being the golden ratio ;)

    Maybe you could proof that this number emerges from the decimal system and using that the golden ratio is the limit of sqrt(1+sqrt(1+sqrt(1+....)))

    But that's just a wild guess :D

  • I was actually thinking it wouldn't be that hard to write some code for this problem as I usually end up writing some C code myself when I'm looking for patterns.

    To be honest I'm more interested how did you end up with 3-sqrt(5). You should do a follow up on that.

  • Just out of curiosity, did you employ a regex backreference to identify the roots that matched?

  • This is why people find maths boring.

  • @jamma246 He says in the video that he's never seen anyone talk about this before. How could this possibly be why people find maths boring if most mathematicians haven't seen it, let alone people such as yourself?

  • 8 works too, 2.828427.......

  • @knightnicholasd like writing comments on YouTube videos to someone you've never met and know nothing about?!

  • @knightnicholasd Lots of things in sciences don't seem to have "functions" at the first glance, but that doesn't mean they're useless! Most people thought the internet and computers would be useless when they were first dreamt up!

  • @knightnicholasd This problem prompted him to learn a programming language. How can learning a programming language not be considered useful?

  • @knightnicholasd

    He focused his energy into learning computer programming because of this, which is going to help in other ways. He's also introducing a new concept to the pure math world, which applied math people will then take as a tool and potentially find it's application. There are many examples where pure math such as topology and number theory have actually found application in nature, so I'm sure somewhere along the line we will find one for this concept.

  • @knightnicholasd You must be an engineer.

  • @Lyrelia Big talk from a waitress.

  • does anybody know the means by which he found that the number approached 3-sqrt(5)

  • I would imagine these grafting numbers are a lot more frequent as you lower the BASE. Is their some kind of relationship between the BASE used and the frequency. Now there's a project to keep you busy!

  • @MelTurpin As you decrease the base the no of usable digits decreases.. so its more likely to find repeating patterns...

  • Looks like top commentors know more then this devoted numberphile

  • 999998 is 999.999 not 999998

  • @Adj19888 Try using a calculator with a higher number of places displayed. Until the 10-billionths (10^-10) spot you will only see 9s after the decimal, where you will find a 4.

  • Also, the resemblance to a Monty Python sketch comes to mind...

  • Yeah. Python coding you good sir, win.

  • Welcome to the Python world, I recommend you to take a look at Numpy!

  • This is my favourite numberphile so far. Intriguing start, simple beginning and rabit-hole type depth. Excellent! (Thank you.)

  • Hitler would like this video. NEIN NEIN NEIN NEIN NEIN NEIN NEIN NEIN NEIN!

  • @RasmusLastname ROFL

  • Python FTW!

  • @pabloso Real men use FORTRAN.

  • Surely you missed 77 off this family, which you had as part of a different family earlier? That way you can multiply by 100 for the first step rather than 1000

  • Or even 8, which would be the first term using your formula (square root 2.8....)

  • Also, have fun generalizing this idea over higher roots :-P

  • Reminds me of how people start to look for messages, codes or patterns in the math constants. Especially the "message spotters" are hilarious because though math may be strict, deterministic and completely predictable (given you've found the sequence generating function) our language is completely arbitrary.

  • This is beautiful!

    Thank you.

  • @numberphile do a video on the number "42" XD

  • At first I thought this was David Tennant.

  • Synchronicity alert. At the exact moment the camera cut to the clock, I looked at the clock on the mantlepiece. Exactly the same time :D

  • Well now I'm glad I was really into Actionscripting and XML when I was younger. Now I just have to skydive.

  • Now that's a great video! :)

  • I'm amazed by the fact that he remembered so much of these numbers.

  • Comment removed

  • no more such numbers?

    well sqrt(100) = 10.0

    *ba dum tsss*

  • @Cr42yguy I should think that we could simple refer to this as a trivial solution.

  • @Cr42yguy I believe he is talking irrational numbers, not perfect squares.

  • I noticed how 3 and 5 are prime so I played around with 5 and 7 and found that the numbers match from the second decimal place of 5-sqrt(7) and the first decimal place of (5-sqrt(7))^2 as far as wolfram alpha allows which is over 3000 decimal places.

  • @AlphaAndOmeg4 Holy F·$% youre right! i only tried a few cases .....and the non prime pairs seemed to not obey the grafting property until i got to(7- sqrt(37))^2 and (7- sqrt(41))^2, anyway im hooked.

  • @AlphaAndOmeg4 I notice that 7-sqrt(11) doesn't work. But, I bet if you switched to a higher base, it would (don't have time to check this myself). Eg. start doing the numbers in base 16 and try 7-sqrt(B), B-sqrt(D) etc.

  • @parkamark base 16 didn't work. But I found that 5-sqrt(x) where x is an integer between 1 and 24 not including 4 and 16 have the same property. After x=25 of course it become negative.

  • find it strange how he writes numbers and dots. decimal point looks like multiplication. 7 looks like 1.

  • The square root of 100 = 10.0 . I just repeated the numbers.

  • wow!!!! very inspiring ....u got great mind..keep it up.

  • I have recently been on an obsession of infinite tetrations. (Like root 2 to the root 2 to the root 2...) Could you guys PLEASE do a video about them? :)

  • Comment removed

  • This video put me to sleep, keep it coming :P

  • After watching the video on narcissistic numbers., I had gone to write the program to find those as well. I completely agree that the ability to use basic programming is incredibly helpful to a mathematician. Maybe not so incredibly deep into the language like graphical user interface design., but at least far enough to do logic and loops.

  • I think this is one of the more fascinating videos so far. Keep them coming, @numberphile!

  • Oh yeah, I've done this before. :p

  • Doesn't 100 work as a graftying number?

    Like, 10.0?

  • hehe very entertaining :0 though i don't want to become a mathematician, tis still awesome finding patterns like that

  • Somebody give this guy a nobel prize.

  • awesome!

  • I've been doing some manual searching (in excel) for cube grafting-numbers (excluding the obvious, i.e 1, 2, 1000, so on, so far I have:

    cuberoot 999 = 9.99666...

    cuberoot 1535 = 11.535492...

    cuberoot 48570 = 36.48570154...

    cuberoot 646409 = 86.4640946...

    these are all of them up to 700000, found manually :)

  • Good morning. You have been in suspension for: 99999 99999.... This courtesy Call is to inform you that all test subjects should vacate the Enrichment Center immediately.

  • Haha! I love the Python plug. I'm doing the facebook hacker cup problems in Python right now. I would also usher anyone to learn it.

  • First thing I thought of when you said you used excel was "this would have been way easier and more thorough in python".

  • It's over NINE THOUSAND.

  • more numbers that work (exactly):

    sqrt(100)=10.0,

    sqrt(10000)=100.00,

    ...

  • This sounds like what I do on my calculator when I'm bored in class :P

  • @numberphile Like a mathematical tautology; ?

  • anyone who finds patterns like this is a beast. ur awesome man

  • this is severe numberphilia.

  • he makes numbers so sexy!

  • is cool.

  • Very interesting! Can't wait for the episode about the golden ratio!

  • Do a video on 142857

  • one day it can be useful for finding the formula for prime numbers (maybe) :)

  • you should do a video on the Laplace transform. it came up in a couple classes, so i had two professors explain it to me, but still the mathematics behind it leaves me with an incredibly blank stare.

  • PYTHON is awkward. Try RUBY.

  • @knugie

    :O

    Blasphemy! Python is less awkward than Ruby!

  • My though is that, since the square root of 100 is 10 and the square root of 10000 is 100, and so on... why wouldn't numbers that are in a close proximity to these numbers have a square root that also begins with the numbers in the original number?

  • @FatLingon

    Let's say you have (x+e)^2, where x is say 10 and e is a very small number. Then the generated decimals come from (x^2+2ex+e^2) which equals (100+20e+e^2). As "e" gets near zero, the decimals of e^2 are negligible and the generated decimals come from 20e, or even simpler 2e.

    So if you wanna double a number, and only have the square function,

    take 10 point some zeros, fill in your number, and square it, and you find the number doubled in the decimals.

  • I used to be good at math..until I took an arrow to the knee.

  • The starting point is very obvious The square root of 100 is obviously 10, so numbers that approach 100 will approach having a square root exactly 1/10 as big.  I do the same sort of thing for fun :D

  • Irrational square roots have completely random patterns. Eventually, you can find anything.

  • @supergsx Maybe you can eventually connect any numbers together, but the way you connect them has meaning. Besides, just because they have an infinite number of digits doesn't mean they're random. You're more likely to see some digits than others. They're also derived, unclear patterns from clear ones. It's like aperiodic tilings - You can take a mundane high-dimensional object and tile it, but when you flatten it into 2D, you can get a never-repeating pattern.

  • @LokiClock You have a good point (and I commented that before watching the rest of the video and hearing about his algorithm), but finding patterns in chaos (such as the digits in irrational numbers) is much like finding patterns in the alignments of the rocks on Mars. Conclusions in that type of data are completely useless.

  • @supergsx It depends on how well they correspond. If you take another photograph with the same alignments it means nothing, unless you're proposing this is a photograph of the same rocks. If you prove it's a photograph of the rocks on Mars, then it doesn't matter how good the film is, because you can come back later and take a better picture. By virtue, you've proven that no matter how good a picture you take it will continue to look like those rocks, in that alignment.

  • There actually ISN'T a Wikipedia page on Grafting numbers. Just saying...

  • a) this channel should be named mathematical curiosities, good channel otherwise.

    b) so what your saying is that you discovered an underling property of rational numbers as expressed by the decimal system? am quite certain this can be used in cryptography.

  • @AlucardNoir

    Throughout all of these videos, I keep trying to convince myself there is some practical application for these numerical properties. So I think what you mentioned as well as computer security systems.

  • Brilliant!

  • he sure can remember a lot of numbers, awesome vid

  • Python is my favourite programming language.

  • Your number 7 and ur brackets look the same, maybe u should put a line throught the 7 like a F (sorta) and be less confusing and cool :D

  • This should have been incredibly boring to watch. Yet it was presented with such enthusiasm, it was completely enjoyable to watch!

    Very good video. Upvoted!

    By the way, you have one helluva good memory for numbers if you could just write them down from your head.

    I wonder if this can be used to break public key cryptography! :D

  • 2-(3-sqrt(5))/2 = golden ratio= 1.6180339887....

    Coincidence? I have no idea.

  • @mungorn we've got some golden ratio stuff coming soon

  • @numberphile Looking forward to it!

  • @numberphile One of my favorite things about the use of the golden ratio would have to be in the formula for the n-th Fibonacci number (derived via linear algebra finding the eigenvalues of the matrix [{1,0}{1,1}]). The original discovery of the number is really fun too (the similar rectangles proof)

  • @mungorn That's how you calculate the golden ratio. You set up G -1 = 1/G and solve and that's what you get.

  • @mungorn that can be found using sequences and series :)

  • @mungorn Here is a better way of presenting it: the golden ratio=(1+sqrt(5))/2 and (3-sqrt(5))/2 ; add both ratios together ,they’ll give you 2 exactly.

  • @mungorn A much quicker way to calculate the golden ratio is:

    0.5 + sqrt(1.25) = 1.6180339887498948482045868343­656

  • @Gogargoat Yes but, theres the 3-sqrt(5) where they talk about in the vid... So there might be a relation with that...

  • @mungorn

    GoldenRatio = (1 + sqrt(5)) / 2

  • @mungorn Or X-1=1/X or x+1=1/x X=1.6180339887... x=0.6180339887... golden ratio is fun..

  • i like how the root of 98 = 9.8994949.. and root of 99 = 9.949 shares the exact same digits for the first few places. when you take out the 9.8 in root of 98

  • Never mind, as he said, thought processes are generally not original....

  • Sq. root 100 is 10.0

    lol

  • @PeachesAndSkeet lol there you go, you found a complete family of grafting numbers is you keep going (sqrt(10000) etc...) yey

  • @EvilTwinkie15 Hahaha I have them memorized too, just like Matt!!

  • What happens when you square root (3 - sqrt(5)) then? :D

  • wouldnt 100 be a grafting number....

  • the numbers... WHAT DO THEY MEAN?

  • I got so excited when I saw the pattern emerging before he mentioned it, then he mentioned it, I cant explain this emotion. Now I want to try it in different bases, would you mind sharing the code you made?

  • When Matt started finding the pattern behind his grafting numbers, I thought to myself, "Damn, this is turning into some Da Vinci Code shit."

  • python += 1

  • Yeah, I'm not jumping out of a plane even if I can.

  • Does sqrt(100) = 10.0 count? That seems like a family of quite base(ic) grafting numbers to consider.