@@snekback. Yes, but I learned it years ago and well understood. I tried to do something interesting with it so I researched hyper roots and logs, tried to differentiate it. When this video came out I thought: what if the topic will be researched more?
no one even knows what x^^0.5 is iirc, but derivitive of x^^n is.. well.. math.stackexchange.com/questions/617009/finding-the-derivative-of-x-uparrow-uparrow-n
IU am 66 yrs old. I earned a MS in Mathematical Physics in 1977. I never heard about Tetration till just now THANK YOU so much!!!!!!!!!!!!!!!!!!!!!!!!!
19midnightsun87 since titration wasn’t coined until the 1980s it would make sense that Cold Logic Crusader would not know it. Heck, no computer of that day could even try to perform the operation.
Also x↑↑↑y = x↑³y is called pentation and other notation to represent it is x [5] y x↑ⁿ⁻² y = x [n] y, where n is the number of operation (ex: sum n=1, multiplication n=2, potentiation n=3, tetration n=4, etc).
Tetration, not to be confused with titration. I wonder how many chemistry students came here looking for curves and then subsequently ran away because of some light mathing. Love the video. Clear and concise. It's clear that your professor chops are strong.
I found a nice recursive formula for the derivative of x↑↑n by setting y = x↑↑n, taking the log on both sides and doing implicit differentiation: d(x↑↑n) = x↑↑n * ( d(x↑↑(n-1)) * log(x) + x↑↑(n-1) / x )
For any function f(x)^g(x), the derivative can be found by adding the power rule to the exponent rule. That is to say d/dx (f(x)^g(x)) = f’(x)*g(x)*f(x)^(g(x)-1)+g’(x)*ln(f(x))*f(x)^g(x)
@@gregsouza7564 I actually proved this in my math class lol it’s not that difficult to derivate f(x)^g(x) and determine this result. In fact if you note that in cases c^a and a^c the constant parts have their half of the equation canceled out due to having a derivative of 0. Therefore the power rule and the exponent rule are both just simplified versions of this general rule
Doh, I didnt include a link! Here they are. 175- th-cam.com/video/EUvFXd1y1Ho/w-d-xo.html 176- th-cam.com/video/Z8I68E7yZeY/w-d-xo.html Sorry about that.
I didn't read through the comments so if someone has already posted the derivative then kudos to them. The form of the derivative of x^^n for n >=4 is: x^^n*x^^n-1*(1/x+x^^n-2*(lnx/x+x^^n-3*(ln^2x/x+x^^n-4*(ln^3x/x+...+x^^2*(ln^(n-3)x/x+ln^(n-2)x+ln^(n-1)x)...) You can prove this by induction. The inductive step is shown by the recursion derivative of x^^n = x^^n*(x^^n-1/x+lnx*d/dx(x^^n-1)) and the base case is that the derivative of x^^4 is x^^4*x^^3*(1/x+x^^2*(lnx/x+ln^2x+ln^3x)). I put the base case in the same form as my answer to show that it's true because yt comments are hard to format and anyways loads of people in the comments did the x^^4 case. Moving on to the induction we have d/dx(x^^n+1)=x^^n+1*x^^n*(1/x+lnx/x^^n*d/dx(x^^n)) from the recursion. Then we plug in the same form from above. d/dx(x^^n)=x^^n*x^^n-1*(1/x+x^^n-2*(lnx/x+x^^n-3*(ln^2x/x+x^^n-4*(ln^3x/x+...+x^^2*(ln^(n-3)x/x+ln^(n-2)x+ln^(n-1)x)...) Therefore: d/dx(x^^n+1)=x^^n+1*x^^n*(1/x+x^^n-1*(lnx/x+x^^n-2*(ln^2x/x+x^^n-3*(ln^3x/x+x^^n-4*(ln^4x/x+...+x^^2*(ln^(n-2)x/x+ln^(n-1)x+ln^(n)x)...)) Done.
Hi, I really like your videos and since you used tetration in this one, I would like to ask you a question that's been on my mind for quite some time now, but for which I could not find a solution yet. What I realized was the following: If you try to complete the natural numbers with respect to subtraction, which is the inverse operation to multiplication, you get the integers. If you complete these with respect to quotients, which are inverse to multiplication, i.e. form the quotient field, you obtain the rational numbers. By completing these wrt. roots of polynomials, i.e. wrt. exponentiation, you obtain the algebraic numbers. But what if you complete these using tetration, i.e. add superroots, superlogs and other solutions of "tetration equations"? E.g. the superroot of 2, i.e. the number x with x^x=2, seems to not be algebraic, so can you form a field extending the rational numbers that is closed wrt. these operations? It can't be a field extension of finite dimension since it is not algebraic, and it also has to be a field, I think, and it should still be countable so it isn't the real numbers, but I could not figure out much more. Also, if you continue this process with pentation and higher order operations (see Knuth's up-arrow notation), you should get other fields extending each other. Since they are a mapping family, you can take the direct limit of those, and get a very big field - are these the whole real numbers? A lot of questions, I know, but I hope someone else already thought about it and figured it out, since it is far beyond my current scope. Anyway, I would be happy about any kind of help.
They should be included in the real numbers, moreover if we keep going towards the higher orders, we could find the transcendental constants (Like π, e) midway through. Or, we might even try to approach infinitation (as in, the infinth order, writing with infinitely many up arrows) and 1) find certain numbers which can't be written without the inverse infinth order. Or, 2) already cover up every real number, thus having no real numbers left, furthermore having found a way to cover up the first uncountable infinity. Or, 3) just get to the transcendental numbers with a clear definition about them.
Hey BPRP. I'm about to graduate with a math degree. I want to tell you, your vids are awesome and I love your pronunciation. Thanks for what you do! You bring me snippets of math I can enjoy when I feel bogged down in technical math i have to learn for a grade (which I'm sure you know can suck the fun out of it). So thanks. Truly.
So I was just learning multivariable calculus and I realized how much simpler this is if you just use the multi variable chain rule on f(x, x^x) where f(y,z)=y^z
For pentation, you have to use the up arrows, presumably because you run out of upper corners to write in after tetration due to the number of upper corners being 2...
bl00dwork OK, I read the "^^" as the notation for what bprp introduced in this video, so e.g. x^x^x would be x^^3. But I have no idea myself about what i^^i could mean... I'm not sure if it even makes sense anyway...
By replacing those "i"es with e^(i*pi/2), i^i^i would be e^((i*pi/2)*e^(i*pi/2)^i) :( and simplifing this we get e^(i*pi/2 * e^(-pi/2)):-| is's still a mess but that's it. if you want to expand it with Euler's formula again, it would be much more "complex":-)
I think you can also do it with implicit differentiation where you take the natural log on both sides. You'll need to repeat for x^^3 so for x^^2 you'll have: x^^2=y XLn(X)=ln(y) Ln(X)+1=Dy/DX •1/y Therefore Dy/DX=x^^2(ln(X)+1)
I think you forgot to divide by x in the second term on the left-hand side of line 3, but yeah I used implicit differentiation and got the right answer, too(as I tend to do first when I don't know how to take a derivative in general) Edit: I don't know why I thought you made a mistake in line 3, but it all looks fine upon rereading it.
So, this notation that blackpenredpen just told, I had actually once accidently discovered it myself. So, when I did a bit of research on values of x^^a for fractional values of 'a' and found out an elegant relation between x^^(1/2) and Lambert - W function. It is: x^^(1/2)=e^(W(x)) or W(x)=ln(x^^(1/2)). I am also working on the derivative of x^^a, which is partially answered by Calyo Delphi in one another comment. I did some research and thought I should share it here.
3 years late but I think you might be wrong because e^^x is (as far as I know) an injective function, so if e^^a = e^^b then a = b. But e^^1 by definition equals e, and if you plug in e into your formula {x^^(1/2)=e^(W(x)}, you get e^^(1/2) = e^W(e), which is still e. This however contradicts e^^x being an injective function, since 1/2 ≠ 1, therefore the formula is incorrect
@@vedantneema Wow thanks for replying early lol. However could there be a misconception from your first line of proof? You see, log_x(x^^n) = x^^(n-1) => log_x(x^^1) = x^^0 x^^1 = x ∀ x (by definition) => log_x(x^^1) = log_x(x) = 1 = x^^0 ∴ x^^0 = 1 ∀ x But in your 1st line of proof, you imply that x = t^^n => t = x^^(1/n) for all n. This says that 2 = t^^n => t = 2^^(1/n). However as n approaches to infinity, t = 2^^(1/n) approaches 2^^0, which from the lemma above shows it approaches 1. This cannot be true though since; 2 = lim_(n -->∞)(t^^n) => 2 = t^t^t^t^... => 2 = t^2 (since t^t^t^... appears in itself) => t = √2 This shows that t = √2 and t = 1, but √2 ≠ 1, making a contradiction
@@fantiscious "x = t^^n => t = x^^(1/n) for all n". No, I only meant it for when n = 2. Also in your last statement you conclude that t can have two solutions for the given equation, one each is obvious in two different but equivalent forms. That stems from the fact that unlike addition, multiplication or exponentiation; tetration is not monotonic. Refer to the graph of x^x within [0, 2](decreasing in [0, 1] and increasing within [1,2]). The graph of y^y=x (== y = x^^1/2) is equally weird. Now the limits within which x^^n is monotonic is something that may be interesting to work out. I speculate it remains the same for n>=1. I'll report back if I find something.
(I'm new in calculus so please don't judge me too badly) A nice rule for the derivative of n^xcould be d/dx n^x=n^x*(n-1)^x...2^x(ln(x)^(n-1)+ln(x)^(n-2)+ 1/x (ln(x)^(n-3)+ln(x)^(n-4) +...+1)). The approach have been to do the derivative for 3^x 4^x and 5^x and I noticed this trend.
For Expert: Tetration n times: Write it as T(x,n) so that T(x,1) = x, T(x,2) = x^x and so on, write its derivative as T'(x,n) Construct a reduction formula, start from n=2, try to solve T'(x,n) for all integers n. (tedious like hell)
Left-upper index is usually used for bottom-order tetration, for right-order hyperoperators arrow notation is basic. Left hyperoperators are rarely used though so it's specified explicitly in that case.
If we have the equation y = x^^n where n is a constant, we can derive a general formula for the derivative using logarithmic differentiation: ln y = ln x^^n x^^n is equal to x^x^x... with n x's. This can also be rewritten as x^(x^^(n-1)). Substituting this in we get: ln y = ln x^(x^^(n-1)) Using the rules of logs, we can move the power to the outside: ln y = x^^(n-1) ln x Let's try differentiating it as is and ignoring the new tetration; maybe we can find a recursive rule by doing this. For the sake of brevity, z = x^^(n-1): ln y = z ln x 1/y dy = ln x dz + z/x dx dy/dx = y(ln x dz/dx + z/x) dy/dx = x^^n (ln x dz/dx + (x^^(n-1))/x) Now we have the derivative in terms of dz/dx, the derivative for the previous value of n. We can use a base case to make this a recursive function. The easy option would be n=1, because when n=1 the function is simply y=x, and the derivative is 1. In order to make the notation simpler, let's define f like this: f[n](x) = x^^n (Picture [n] to mean n in the subscript, obviously I can't type that easily on a youtube comment) Now we can write a recursive definition of f[n]'(x): f[1]'(x) = 1 f[n]'(x) = f[n](x) (ln x f[n-1]'(x) + f[n-1](x)/x) Let's expand the second part: f[n]'(x) = f[n](x) ln x f[n-1]'(x) + f[n](x) f[n-1](x)/x Now let's expand f[n-1]'(x) with the previous iteration of the function, to see if we can find a pattern to make into an explicit rule: f[n]'(x) = f[n](x) ln x (f[n-1](x) ln x f[n-2]'(x) + f[n-1](x) f[n-2](x)/x) + f[n](x) f[n-1](x)/x f[n]'(x) = f[n](x) ln x f[n-1](x) ln x f[n-2]'(x) + f[n](x) ln x f[n-1](x) f[n-2](x)/x + f[n](x) f[n-1](x)/x f[n]'(x) = f[n](x) f[n-1](x) (ln x)^2 f[n-2]'(x) + ln x f[n](x) f[n-1](x) f[n-2](x)/x + f[n](x) f[n-1](x)/x Now let's expand it one more time: f[n]'(x) = f[n](x) f[n-1](x) (ln x)^2 (f[n-2](x) ln x f[n-3]'(x) + f[n-2](x) f[n-3](x)/x) + ln x f[n](x) f[n-1](x) f[n-2](x)/x + f[n](x) f[n-1](x)/x Now we can see what would happen if we expand this again: 1. The first term gets another ln x and the next f function (n, n-1, n-2, ...) 2. A new term with one higher degree of ln x than the last new term and the next f function is added Now we can create an explicit rule for this derivative: f[n]'(x) = f[n](x) f[n-1](x) ... f[2](x) (ln x)^(n-1) + (f[n](x) f[n-1](x) + ln x f[n](x) f[n-1](x) f[n-2](x) + ... + (ln x)^(n-2) (f[n](x) f[n-1](x) f[n-2](x) ... f[1](x)))/x f[n]'(x) = Π[k=2→n](f[k](x)) (ln x)^(n-1) + Σ[k=1→n-1]((ln x)^(k - 1) Π[h=0→k](f[n-h](x)))/x (for all n ≥ 2) This is a very long rule, however, it is practical to use. Let's try using it for n=3 to see if we get the same result as in the video: f[3]'(x) = Π[k=2→3](f[k](x)) (ln x)^(2) + Σ[k=1→2]((ln x)^(k - 1) Π[h=0→k](f[3-h](x)))/x f[3]'(x) = f[2](x) f[3](x) (ln x)^2 + ((ln x)^0 f[3](x) f[2](x) + (ln x)^1 f[3](x) f[2](x) f[1](x))/x f[3]'(x) = f[2](x) f[3](x) ((ln x)^2 + 1/x + ln x f[1](x)/x) f[3]'(x) = x^^2 x^^3 ((ln x)^2 + 1/x + ln x x^^1/x) f[3]'(x) = x^^2 x^^3 ((ln x)^2 + 1/x + x ln x/x) f[3]'(x) = x^^2 x^^3 (1/x + ln x + (ln x)^2) Indeed, this is exactly the same as the result in the video, which means that this equation is most likely correct. Here is the equation written in LaTeX for anyone who wants it: www.codecogs.com/eqnedit.php?latex=\frac{d}{dx}(^nx)&space;=&space;\left\{&space;\begin{array}{ll}&space;(\ln&space;x)^{n-1}\displaystyle&space;\prod_{k=2}^{n}(^kx)&space;+&space;\frac{1}{x}\displaystyle&space;\sum_{k=1}^{n-1}((\ln&space;x)^{k-1}\displaystyle&space;\prod_{h=0}^{k}&space;(^{n-h}x))&space;&&space;\quad&space;n&space;>&space;1&space;\\&space;1&space;&&space;\quad&space;n&space;=&space;1&space;\end{array}&space; ight%2E
@@ozanbayrak562 Using super Logarithm (inverse of Tetration) By definition sLog2 (2^^3) = 3 NOTE: "sLog" is a notation for super Logarithm. Like how Logarithm cancels the base leaving the exponent ex. Log2 (2^3) = 3 super Logarithm does the same with Tetration leaving the super power. We can use super Logarithm to solve non integer super powers since super Logarithm is repeated Logarithm by definition. Let's let sLog2 (16) = 3+x Where 0 ≤ x < 1 (represents a 0 or decimal) sLog2 (2^^3) = sLog2 (2^2^2) => Log2(2^2^2) = 2^2 => Log2(2^2) = 2 =>Log2(2) = 1 At this point we've taken three logs representing our integer part of the solution (given by the fact that the answer is equal to 1). We just take log again for the decimal x (the remainder of 2's that we need.) Log2 (1) = 0 Thus sLog2 (16) = 3+0 = 3 Well let's look at what happens when we go backwards through the same process to see what happens to the remainder. Log2 (Log2 (Log2 (Log2 (16)))) = 0 Log2 (Log2 (Log2 (16))) = 2^0 Log2 (Log2 (16)) = 2^2^0 Log2 (16) = 2^2^2^0 16 = 2^2^2^2^0 = 2^2^2 = 2^^(3+0) The remainder adds an extra '2' to the top of the power tower and the additional 2 is raised to the power of the remainder For 0 ≤ x ≤ 1 By definition sLog a(a^^3+x) => a^a^a^a^x By definition of Tetration a^^3+x = a^a^^2+x = a^a^a^^1+x = a^a^a^a^^x a^a^a^a^^x = a^a^a^a^x a^a^a^^x = a^a^a^x a^a^^x = a^a^x a^^x = a^x by definition for 0 ≤ x ≤ 1 so 2^^1.5 = 2^√2 ≈ 2.6651441427
We still have no clear way of defining how to even calculate non-integer raised tetration, so good luck with that! I mean what would 2^^1.5 even mean?! I've looked into this quite a bit and there seems to be no conclusive answer.
@@DoctorT144 Using super Logarithm (inverse of Tetration) By definition sLog2 (2^^3) = 3 NOTE: "sLog" is a notation for super Logarithm. Like how Logarithm cancels the base leaving the exponent ex. Log2 (2^3) = 3 super Logarithm does the same with Tetration leaving the super power. We can use super Logarithm to solve non integer super powers since super Logarithm is repeated Logarithm by definition. Let's let sLog2 (16) = 3+x Where 0 ≤ x < 1 (represents a 0 or decimal) sLog2 (2^^3) = sLog2 (2^2^2) => Log2(2^2^2) = 2^2 => Log2(2^2) = 2 =>Log2(2) = 1 At this point we've taken three logs representing our integer part of the solution (given by the fact that the answer is equal to 1). We just take log again for the decimal x (the remainder of 2's that we need.) Log2 (1) = 0 Thus sLog2 (16) = 3+0 = 3 Well let's look at what happens when we go backwards through the same process to see what happens to the remainder. Log2 (Log2 (Log2 (Log2 (16)))) = 0 Log2 (Log2 (Log2 (16))) = 2^0 Log2 (Log2 (16)) = 2^2^0 Log2 (16) = 2^2^2^0 16 = 2^2^2^2^0 = 2^2^2 = 2^^(3+0) The remainder adds an extra '2' to the top of the power tower and the additional 2 is raised to the power of the remainder For 0 ≤ x ≤ 1 By definition sLog a(a^^3+x) => a^a^a^a^x By definition of Tetration a^^3+x = a^a^^2+x = a^a^a^^1+x = a^a^a^a^^x a^a^a^a^^x = a^a^a^a^x a^a^a^^x = a^a^a^x a^a^^x = a^a^x a^^x = a^x by definition for 0 ≤ x ≤ 1 2^^1.5 = 2^2^0.5 = 2^√2 ≈ 2.6651441427...
I did get the generalisation for the tetration of x to any natural index(excluding 1).Well,it goes like this, Let x(n) be the tetration of x to a natural number n * denotes the product of the tetrations(pi-product function)[*(k=r to k=n) x(k)=x(r)x(r+1)x(r+2)........x(n) # denotes summation(sigma function) [#(r=1 to r=n) x(r)=x(1)+x(2)+.........x(n)] Then, d[x(n)]/d(x)=(1/x)[*(k=1 to k=n)x(k)(ln(x))^n-1 + #(r=1 to r=n-1)[*(k=r to k=n)x(k)[(ln(x))]^n-r-1]] Where #(r=1 to r=n-1)*(k=1 to k=n) denotes the summation of the products. Put it on paper for easy understanding.Do note it is only valid for natural number greater than or equal to 2. Just put a reply if I had gone wrong somewhere.
For the first deriv of nth tetration of x, namely {^{n}x}I get: (1/x)* \sum_{j=0}^{n-1}(lnx)^{n-j-1}* \prod_{i=j}^{n} {^{i}x} Which can be proved by induction from the recurrence relation. Also we need to complete the pattern for the first and zeroth tetration of x
i can't believe that my friend who have never seen tetration or heard about thought of this concept and wrote it in the same annotation of this and chose to call it superpostion .then he sarted studyng it as a functon and he got some pretty cool stuff .bt he was stuck on a problem.while searching the net for a solution he fond the same concept in the name of tetration and it shook us how similar his invention is to it .
@dolev goaz: Since we're talking mathematics here, I sincerely wonder: *"do you have proof for your two claims?"* (The claims I mean: 1: "you can't integrate x^x" and 2: "so you can't integrate x^(x^x)" ) For x>0 both functions (i.e. exp(ln(x)*exp(ln(x))) and exp(ln(x)*exp(ln(x)*exp(ln(x)))) ) are continuous and differentiable, so why wouldn't you be able to integrate them?
Instead of plugging in exp(xln(x)) each and every time to replace x^x, why not just use the power rule for differentiating the function power of another function, thus if y = u^v, then dy = vu^(v-1)du + (u^v)ln(u)dv So if y = x^x, then dy = xx^(x-1)dx = x^xln(x)dx = x^xdx = x^xln(x)dx = x^x[1 + ln(x)]dx
Tetration! Knuth Arrow Notation! #YAY Note: We can use a reduction formula for the derivative of x^^n d/dx (x^^n) = d/dx (e^((x^^(n-1)ln(x)))) = (x^^n)((x^^(n-1))/x+ln(x) d/dx (x^^(n-1))) Or saying D(n) = (x^^n)((x^^(n-1))/x+ln(x)*D(n-1)))
I think the proper way of pronouncing tetration would be „x tetrated to n“ or „x tetrated to the n-th power“. And btw, just as we write x^n for „x to the power of n“, i think we also could simply write x^^n for „x tetrated to n“.
This is what I got, though I’m not sure if correct! It’s technically an explicit function 😂 d(x || n)/dx = E(k=1 to n) [ {((ln(x))^(n-k))/x} * P(i=k-1 to n) [x || i] ] Notation E ~ the sigma summation function P ~ the sigma pin function || ~ the double arrow function Domain: For n within positive integers and assuming that x || 0 = 1
The rule is x↑↑n = (x↑↑n) * (d/dx ( x↑↑(n-1) * ln(x) + x↑↑(n-1) *(1/x)) Thats the best you can do its basically the original function multiplied by the product rule of x↑↑(n-1) and ln(x)
La solución a lo del final sería algo así como: d/dx(x↑↑n)=fact(a=1 to n-1, x↑↑a)*sum(b=0 to n-1,(ln(x))^b) No es tan simple como la otra, pero creo que sirve, no? En fin, no lo he comprobado. The solution to the final stuff would be something like: d/dx(x↑↑n)=fact(a=1 to n-1, x↑↑a)*sum(b=0 to n-1,(ln(x))^b) It's not so simple as the other one, but I think it works, didn't it? Well, I haven't confirmed it.
@Gerben van Straaten if i'm not mistaken this function is also defined for negative rational numbers that can be written as -a/b where b is an odd integer, however the fact that it would not be continuous still stands
To answer the last question in the video, just from generalization of the result, I’m guessing: d/dx (x^^n) = x^^(n-1).x^^n.(1/x + ln(x) + ln(x)^(n-1))
I was also thinking of using substitution and logarithmic differentiation. So solve for x^u, where u= x^x. Use logarithmic differentiation for u first, then after finding u’, use it to find x^u. I think this is long and messy though.
One thing my Maths Sir once told about differentiating x^x is that "first differentiate it as constant^variable then differentiate it as variable^constant and add them that is the derivative of x^x" idk if it works for any tetration
Funny is that there is a rule that alows us to derivate anything no matter how tetrated or stacked it is. Basically instead of doing what he did you can derivate the whole thing a function at a time from outside to inside and write the product of those partial results
Solve for this; Cow tied at the corner of a circular field is able to graze 3/4th of the field. What is ratio of length of its rope and radius of the field?
I got the answer... yeah, I watched the video quite late, but I found the answer all on my own, and then watched the rest of the video, I also got the answer to that last question you asked 😄( it might not seem like a big deal, but I am in 10th grade 😅😅😅 ) it was fun doing it... keep posting such videos 👍🏻👍🏻👍🏻
ok so i just did probably some of the most ugly math ever and found a general form for the derivative of x^^n. so ill try to write it out here for anyone who cares: I'll start out with just the explicit equation, and then explain the notation. d/dx [x^^n] = (1/x * Σ [ t!(n,k) * (ln(x))^(n-1-k) ] ) + t!(n,2)*( (ln(x))^(n-1) + (ln(x))^(n-2) ), where Σ is from k=2 to k=n-1 first of all, x^^n means x tetrated to the n, or ⁿx, hopefully that was clear ok so, the t!(x,y) is a notation i made up that stands for "tetration factorial", where it multiplies many x^^(something), where (x,y) is the range of the (something) in question for example: t!(4,1) = (x^^4)*(x^^3)*(x^^2)*(x^^1) or: t!(5,3) = (x^^5)*(x^^4)*(x^^3) technically, ive only tested this for the derivatives of x^^5, x^^6, and x^^7, but yeah. do with this information what you will. I know its probably the most hideous way of expressing this but whatever. edit: heres a desmos graph that proves that it works for x^^7 if youd like to see www.desmos.com/calculator/ufqbmbhbzs
Since multiplying a number by a non-integer number and taking a number to a non-integer power are both possible, is tetrating a number to an non-integer number possible? Is it also possible for the other infinitely many arithmetic operations?
y = x^f(x) -> log(y) = f(x) log(x) -> y'/y = f'(x) log(x) + f(x)/x -> y' = y (f'(x) log(x) + f(x)/x) this is the derivation of the recursive formula below
Tetration is the first step to big numbers... The goes pentation, then hexation, then Knut's notation, then Conway's notation, and then the fast-growing hierarchy, which will bring you very far away from what is considered 'normal'.
I was unable to come up with an explicit formula for (d(x↑↑n)/dx), but I was able to come up with a recursive formula (I hope). If someone could tell me if the formula is correct or how to make it explicit, that would be fantastic. Formula: d(x↑↑n)/dx = (x↑↑n)(ln(x)(d(x↑↑(n-1))/dx)+((x↑↑n)/x))
I would like to know if tetration has properties. like, if it was a x^2 * x^3 right there, you would be able to add the exponents, but it dosn't seem to be the case with tetration
I found the formula, bear with me d/dx x||n (I cant draw arrows lol) is equal to [Π (i goes from 2 to n) x||i] * [(Σ (a goes from 0 to n-1) (lnx)^a ÷ Π (q goes from 0 to n-a-2) x||q ] Yeah it's kinda lengthy but try to write it down on a paper and check, it works
I found a quite complicated formula for general integration of tetration functions in wikipedia, here: en.wikipedia.org/wiki/List_of_integrals_of_exponential_functions It is given as exactly 20th example from the top; the formula involves incomplete gamma function and some parameters (n, m, i, j), but it's still not quite clear to me. I have just subscribed your channel :-) Could you devote some time to this formula in one of your future lectures? I mean, how it has been found and how it can be used In practice? Anything that can make it more familiar and clear?
@@nishilsheth9076 You know, derivatives can be calculated using quite simple set of a few rules. They include: a formula for derivative of the product (1), derivative of the quotient (2), derivative of a composite function (3), derivative of a power (4), and derivative of an inverse function (5). So here you are: (1) [f(x)*g(x)]' = f'(x)*g(x) + g'(x)*f(x), or - more simply, (fg)' = f'g + g'f (2) [f(x)/g(x)]' = [f'(x)*g(x) - g'(x)*f(x)] / [g(x)]^2, more simply (f/g)' = (f'g - g'f)/g^2 (3) when u(x) = f(g(x)), u'(x) = g'(x) * f'(u), for example [ln(sinx)]' = (1/sinx)*(-cosx) etc. (4) [f(x)^g(x)]' = [f(x)^g(x)] * {[g'(x) * ln f(x)] + [g(x)/f(x) * f'(x)]}, or more simply: (f^g)' = (f^g) * (g'* lnf + g*f'/f ) - which is an applied formula for derivative of a composite function after making an identity transformation: f^g = e^(g*lnf) (5) when y=f(x) and x=g(y), g'(y)=1/f'(x). For example when x=y^y and y=ssrt(x), (y^y)* (1+lny) = 1/[ssrt(x)]', which yields [ssrt(x)]' = 1/[y^y * (1+ lny)], so: when f(x) = ssrt(x), f'(x) = 1/ [x + ln(ssrt(x)], since when y= ssrt(x), y^y = x by definition When using these five formulas, we can determine a derivative of any function expressed with a finite formula.
I'll try it. So x^(x^x) = y, e^(x^x)ln(x) = y, e^e^(ln(ln(x))+xln(x))=y Derivative is x^(x^x) * (x^x)ln(x) * (1/xln(x) + ln(x) + 1) Which can be simplified to x^(x^x +x) *(1/x + ln(x)^2 + ln(x))
d(e↑x)/dx = e↑x; one of the reasons e is a special number for calculus x = e↑(ln x); an identity that synergizes well with the previous rule d(f(g(x)))/dx = f'(g(x)g'(x); the chain rule of calculus. x↑x = e↑(x ln x) Finding the derivative of x ln x requires the product rule: d[f(x)g(x)]/dx = f'(x)g(x) + f(x)g'(x) 1* ln x + x * 1/x ln x + 1 d(x↑x)/dx = (1 + ln x) e↑(x ln x) which simplifies to (1 + ln x) x↑x
Man, I've been studying this relation for hours, and something seems to pop up. First, I've found a reduction formula. Second, I've seen this behavior pop up: The derivative of x^x is x^x(1+ln(x)), normal, right? x^x^x is x^x^x*x^x(1/x+ln(x)+ln^2(x)), the last two are normal. x^x^x^x is x^x^x^x*x^x^x*x^x(1/(x^x*x)+ln(x)/x+ln^2(x)+ln^3(x)), we see the behavior 1 over x^x*x, x, 1, 1 x^^5 is with behaviour x^x^x*x^x*x, x^x*x, x, 1, 1 So it's for now the best to study why this behavior happens and then we'll have a formula to calculate d/dx (x^^n)
I think the behavior happens because when we apply my reduction formula, we have new tetrations of x which add on to the 1/x, and the previous terms are preserved. D(3)=(x^^3)((x^^2)/x+ln(x)*x^x(1+ln(x))) = x^^3*x^^2(1/x+ln(x)+ln^2(x)) The cancellation adds 1/x D(4)=(x^^4)((x^^3)/x+ln(x)*x^^3*x^^2(1/x+ln(x)+ln^2(x))) = x^^4*x^^3(1/x+ln(x)^x^^2(1/x+ln(x)+ln^2(x))) But then we have to divide x^^2 which adds to the 1/x term, so it becomes x^x*x = x^^4*x^^3*x^^2(1/x^x*x+ln(x)(1/x+ln(x)+ln^2(x))) And the other terms are preserved. So you see what I mean?
Why is my comment not showing up?!? Anyway, in the comment, I wrote this: "We can use a reduction formula for the derivative of x^^n d/dx (x^^n) = d/dx (e^((x^^(n-1)ln(x)))) = (x^^n)((x^^(n-1))/x+ln(x) d/dx (x^^(n-1))) Or saying D(n) = (x^^n)((x^^(n-1))/x+ln(x)*D(n-1)))"
Solve x^x^x=2? Check out here: th-cam.com/video/ef-TSTg-2sI/w-d-xo.html
What if
d(3↑↑X)/dx?
Professor: "On exam, I will test whether you know mathematical induction"
Exam: "Find derivative of n-th tetration of x." 🤣🤣
sqrt(2)
@@MrAleksander59 The notation would be hard enough already
@@snekback. Yes, but I learned it years ago and well understood. I tried to do something interesting with it so I researched hyper roots and logs, tried to differentiate it. When this video came out I thought: what if the topic will be researched more?
“i don’t know how to integrate this so don’t ask me”
we found his kryptonite
Lmaooo
no one even knows what x^^0.5 is iirc, but derivitive of x^^n is.. well.. math.stackexchange.com/questions/617009/finding-the-derivative-of-x-uparrow-uparrow-n
@@incription scary
Wha...? Well now, I don't feel so dumb.
it would just be x^(x^(x))+1)/x^x +1) +c
Now integrate it
No, it's hard!
😂😂😂😂😂
X^x^x + c
If you integrate x^(x^x) (i.e. 3rd tetration of x), then there’s no answer since it’s non-elementary. If you integrate x^3, it would be x^4/4+C.
@@justabunga1 What does elementary mean?
IU am 66 yrs old. I earned a MS in Mathematical Physics in 1977. I never heard about Tetration till just now THANK YOU so much!!!!!!!!!!!!!!!!!!!!!!!!!
Thank you!!
Hm, seems like the education back then was pretty shitty if you didn't hear about that. Where did you earn your degree?
@@19midnightsun87 Grow up.
19midnightsun87 since titration wasn’t coined until the 1980s it would make sense that Cold Logic Crusader would not know it. Heck, no computer of that day could even try to perform the operation.
Ah, I see! Very interesting. Thank you for the info.
"Im just gonna put this in the thumbnail to make a little clickbait"
Transparency
Sree Kommalapati lolllll and then i changed
If you'd put d/dx (x³) for my final exam, you'd be my favourite teacher!
If that appears in your final exam i assume you dont need a teacher
at least solve it by base principles
Let me give you one@@1001-t4z
get vaxxed! You need at least 3 or you won't pass the test
@suyunbek1399 no sir. I will never remember the limits definition of derivatives. Power rule all the way!
For those who'd like to do more research on ⁿa, the notation is called tetration.
Also x↑↑↑y = x↑³y is called pentation and other notation to represent it is x [5] y
x↑ⁿ⁻² y = x [n] y, where n is the number of operation (ex: sum n=1, multiplication n=2, potentiation n=3, tetration n=4, etc).
Thanks! I didn't know nothing about this.
Hows the thing called you need for g_64?
ahh interesting, thanks :-)
@Connor Gaughan if thats actually the case.. how boring
X: This isn't even my final form!
Is this a reference to Dragon Ball? Cell's final form or something.
@@blackpenredpen shaggy
@@blackpenredpen Freeza
Tetration, not to be confused with titration. I wonder how many chemistry students came here looking for curves and then subsequently ran away because of some light mathing. Love the video. Clear and concise. It's clear that your professor chops are strong.
none what so ever
@@amineaboutalib Prove it
@@amineaboutalib *whatsoever
Im a chemist and I came for the math :)
Lmao 🤣
I found a nice recursive formula for the derivative of x↑↑n by setting y = x↑↑n, taking the log on both sides and doing implicit differentiation:
d(x↑↑n) = x↑↑n * ( d(x↑↑(n-1)) * log(x) + x↑↑(n-1) / x )
I'm not sure this is that clean and nice.
@@hugoburton5222 well it's recursive, so it probably can't get cleaner than this
I think it is simpler ;-) -> d(x↑↑n) = x↑↑n * d(x↑↑[n-1] * ln[x])
@ you mean n -> to |R? the question would be, what it would mean......
Now do d/dx(x↑↑x)
For any function f(x)^g(x), the derivative can be found by adding the power rule to the exponent rule. That is to say d/dx (f(x)^g(x)) = f’(x)*g(x)*f(x)^(g(x)-1)+g’(x)*ln(f(x))*f(x)^g(x)
This is a great simplifying formula to show properties in elements in hypercubes :)
This is stolen. You clearly don't know how this is derived. You don't just magically add the formulas together and get this.
@@gregsouza7564 I actually proved this in my math class lol it’s not that difficult to derivate f(x)^g(x) and determine this result. In fact if you note that in cases c^a and a^c the constant parts have their half of the equation canceled out due to having a derivative of 0. Therefore the power rule and the exponent rule are both just simplified versions of this general rule
@@gregsouza7564 lmao whatt? What do you mean "this is stolen". goofy
If you dont derive all math yourself you basically stole it
Good use of the Chen Lu; 70/10
Yay!!!
"Chen Lu" The Goddess of Derivatives
Chain rule haha i died
Lmao
LMAO
Lol I always thought it was "chair rule" because that's how my professor pronounced it. Awesome teacher tho
just wait till you see the "prada lu" the mightiest of all!
You taught me so many things like double factorials, hyperpowers... I never thought such things exist. Well done!
I have never seen this notation before!
You might find this interesting. (Watch 175 and 176).
It starts off a little "hokey", but does get pretty interesting (and introduces that notation)
Silas Rodrigues Its very uncommon. Only really used to describe for extremely large numbers, like Graham's numbee
Doh, I didnt include a link! Here they are.
175- th-cam.com/video/EUvFXd1y1Ho/w-d-xo.html
176- th-cam.com/video/Z8I68E7yZeY/w-d-xo.html
Sorry about that.
By that you mean "I have never seen Numberphile before"?
Me too! Today I learned a brand new thing! >:3
Love your videos: I haven't studied Maths for a long time, and neither do I teach it, but these make difficult problems so easy to follow.
: ))))
I didn't read through the comments so if someone has already posted the derivative then kudos to them.
The form of the derivative of x^^n for n >=4 is:
x^^n*x^^n-1*(1/x+x^^n-2*(lnx/x+x^^n-3*(ln^2x/x+x^^n-4*(ln^3x/x+...+x^^2*(ln^(n-3)x/x+ln^(n-2)x+ln^(n-1)x)...)
You can prove this by induction. The inductive step is shown by the recursion derivative of x^^n = x^^n*(x^^n-1/x+lnx*d/dx(x^^n-1)) and the base case is that the derivative of x^^4 is x^^4*x^^3*(1/x+x^^2*(lnx/x+ln^2x+ln^3x)).
I put the base case in the same form as my answer to show that it's true because yt comments are hard to format and anyways loads of people in the comments did the x^^4 case.
Moving on to the induction we have d/dx(x^^n+1)=x^^n+1*x^^n*(1/x+lnx/x^^n*d/dx(x^^n)) from the recursion. Then we plug in the same form from above.
d/dx(x^^n)=x^^n*x^^n-1*(1/x+x^^n-2*(lnx/x+x^^n-3*(ln^2x/x+x^^n-4*(ln^3x/x+...+x^^2*(ln^(n-3)x/x+ln^(n-2)x+ln^(n-1)x)...)
Therefore:
d/dx(x^^n+1)=x^^n+1*x^^n*(1/x+x^^n-1*(lnx/x+x^^n-2*(ln^2x/x+x^^n-3*(ln^3x/x+x^^n-4*(ln^4x/x+...+x^^2*(ln^(n-2)x/x+ln^(n-1)x+ln^(n)x)...))
Done.
Recursion? nah. Hold my beer: \frac{d}{dx}{^{n}x} = \frac{1}{x}\sum_{k=1}^n\left ( \ln^{k-1}(x) \prod_{i=0}^k {^{n-i}x}
ight )
It's so amazing to see a (mostly) friendly community of people who like math as much as me (:
Hi, I really like your videos and since you used tetration in this one, I would like to ask you a question that's been on my mind for quite some time now, but for which I could not find a solution yet.
What I realized was the following: If you try to complete the natural numbers with respect to subtraction, which is the inverse operation to multiplication, you get the integers.
If you complete these with respect to quotients, which are inverse to multiplication, i.e. form the quotient field, you obtain the rational numbers.
By completing these wrt. roots of polynomials, i.e. wrt. exponentiation, you obtain the algebraic numbers.
But what if you complete these using tetration, i.e. add superroots, superlogs and other solutions of "tetration equations"? E.g. the superroot of 2, i.e. the number x with x^x=2, seems to not be algebraic, so can you form a field extending the rational numbers that is closed wrt. these operations? It can't be a field extension of finite dimension since it is not algebraic, and it also has to be a field, I think, and it should still be countable so it isn't the real numbers, but I could not figure out much more.
Also, if you continue this process with pentation and higher order operations (see Knuth's up-arrow notation), you should get other fields extending each other. Since they are a mapping family, you can take the direct limit of those, and get a very big field - are these the whole real numbers? A lot of questions, I know, but I hope someone else already thought about it and figured it out, since it is far beyond my current scope. Anyway, I would be happy about any kind of help.
have you got an answer to this yet?? sounds insane i wanna know
They should be included in the real numbers, moreover if we keep going towards the higher orders, we could find the transcendental constants (Like π, e) midway through. Or, we might even try to approach infinitation (as in, the infinth order, writing with infinitely many up arrows) and
1) find certain numbers which can't be written without the inverse infinth order. Or,
2) already cover up every real number, thus having no real numbers left, furthermore having found a way to cover up the first uncountable infinity. Or,
3) just get to the transcendental numbers with a clear definition about them.
New math lore
In the ending result you also could simplify: x^(x^x)*x^x = x^(x^x+x), but this depends on which notation you prefer. ;-)
My mans ADMITTED he was gonna put tetration of x in video for a clickbait. HahAHA
Tetration is read as "the nth tetration of a"
x^^3, x^x^x, "the third tetration of x"
or 'x tetrated to 3'?
Hey BPRP. I'm about to graduate with a math degree. I want to tell you, your vids are awesome and I love your pronunciation. Thanks for what you do! You bring me snippets of math I can enjoy when I feel bogged down in technical math i have to learn for a grade (which I'm sure you know can suck the fun out of it). So thanks. Truly.
Thank you!! I am glad to hear this!! Best of luck to you in everything you do.
So I was just learning multivariable calculus and I realized how much simpler this is if you just use the multi variable chain rule on f(x, x^x) where f(y,z)=y^z
For pentation, you have to use the up arrows, presumably because you run out of upper corners to write in after tetration due to the number of upper corners being 2...
I love learning new notations! #YAY
Your ball is getting smaller and smaller. 😁😁😁
@Venky Wank good one XD
Which one of the ~two~...three...?
balls
by ball u mean mic or real balls
@@trustnobody90....the mic
What is i^^i?
Flamingpaper e^(-pi/2)
bl00dwork No, this would be for i^i...
novidsonmychannel justcommenting but I thought that ¡^^¡ would be ¡^¡ since i^^^i would be i^i^i
bl00dwork OK, I read the "^^" as the notation for what bprp introduced in this video, so e.g. x^x^x would be x^^3. But I have no idea myself about what i^^i could mean... I'm not sure if it even makes sense anyway...
By replacing those "i"es with e^(i*pi/2), i^i^i would be e^((i*pi/2)*e^(i*pi/2)^i) :(
and simplifing this we get e^(i*pi/2 * e^(-pi/2)):-|
is's still a mess but that's it.
if you want to expand it with Euler's formula again, it would be much more "complex":-)
I think you can also do it with implicit differentiation where you take the natural log on both sides. You'll need to repeat for x^^3 so for x^^2 you'll have:
x^^2=y
XLn(X)=ln(y)
Ln(X)+1=Dy/DX •1/y
Therefore
Dy/DX=x^^2(ln(X)+1)
I think you forgot to divide by x in the second term on the left-hand side of line 3, but yeah I used implicit differentiation and got the right answer, too(as I tend to do first when I don't know how to take a derivative in general)
Edit: I don't know why I thought you made a mistake in line 3, but it all looks fine upon rereading it.
I'm so pleased I can follow these. You are so fun to watch. Thank you for sharing the awesome math
Differentiation made easier by taking logs of both sides, twice.
taking y = x^x^x ; logy = x^x logx
Taking logs of both sides again, log(logy) = log(x^x) + log(logx)
ie log(logy) = x logx + log(logx)
Now differaentiating both sides with respect to x,
(1/logy)(1/y) dy/dx = x(1/x) + logx + (1/logx)(1/x)
Hence dy/dx = y logy [ 1 + logx + (1/xlogx) ]
= x^x^x . x^x logx [ 1 + logx + (1/xlogx) ]
So, this notation that blackpenredpen just told, I had actually once accidently discovered it myself.
So, when I did a bit of research on values of x^^a for fractional values of 'a' and found out an elegant relation between x^^(1/2) and Lambert - W function. It is:
x^^(1/2)=e^(W(x))
or
W(x)=ln(x^^(1/2)).
I am also working on the derivative of x^^a, which is partially answered by Calyo Delphi in one another comment.
I did some research and thought I should share it here.
3 years late but I think you might be wrong because e^^x is (as far as I know) an injective function, so if e^^a = e^^b then a = b. But e^^1 by definition equals e, and if you plug in e into your formula {x^^(1/2)=e^(W(x)}, you get e^^(1/2) = e^W(e), which is still e. This however contradicts e^^x being an injective function, since 1/2 ≠ 1, therefore the formula is incorrect
@@fantiscious yup you're right. i prob miscalculated then or sth. the actual relation was
(e^x)^^1/2 = e^W(x)
Proof:
x = t^t => t = x^^1/2 -- 1.
x = te^t => t = W(x) -- 2.
=> e^(x) = e^(te^t)
=> e^x = (e^t)^(e^t)
=> e^t = (e^x)^^1/2 = e^(W(x)) (from 1.)
=> (e^x)^^1/2 = e^W(x)
@@vedantneema Wow thanks for replying early lol. However could there be a misconception from your first line of proof? You see,
log_x(x^^n) = x^^(n-1)
=> log_x(x^^1) = x^^0
x^^1 = x ∀ x (by definition)
=> log_x(x^^1) = log_x(x) = 1 = x^^0
∴ x^^0 = 1 ∀ x
But in your 1st line of proof, you imply that x = t^^n => t = x^^(1/n) for all n. This says that 2 = t^^n => t = 2^^(1/n).
However as n approaches to infinity, t = 2^^(1/n) approaches 2^^0, which from the lemma above shows it approaches 1. This cannot be true though since;
2 = lim_(n -->∞)(t^^n)
=> 2 = t^t^t^t^...
=> 2 = t^2 (since t^t^t^... appears in itself)
=> t = √2
This shows that t = √2 and t = 1, but √2 ≠ 1, making a contradiction
@@fantiscious "x = t^^n => t = x^^(1/n) for all n". No, I only meant it for when n = 2. Also in your last statement you conclude that t can have two solutions for the given equation, one each is obvious in two different but equivalent forms. That stems from the fact that unlike addition, multiplication or exponentiation; tetration is not monotonic. Refer to the graph of x^x within [0, 2](decreasing in [0, 1] and increasing within [1,2]). The graph of y^y=x (== y = x^^1/2) is equally weird.
Now the limits within which x^^n is monotonic is something that may be interesting to work out. I speculate it remains the same for n>=1. I'll report back if I find something.
the 10 mins I spent here was worthier than my existence
(I'm new in calculus so please don't judge me too badly)
A nice rule for the derivative of n^xcould be d/dx n^x=n^x*(n-1)^x...2^x(ln(x)^(n-1)+ln(x)^(n-2)+ 1/x (ln(x)^(n-3)+ln(x)^(n-4) +...+1)).
The approach have been to do the derivative for 3^x 4^x and 5^x and I noticed this trend.
Alternatively, you can rewrite the given equation as: ln(ln(y)) = ln(ln(x)) + x*ln(x), and then try differentiating the equation in this form.
For Expert:
Tetration n times:
Write it as T(x,n) so that T(x,1) = x, T(x,2) = x^x and so on, write its derivative as T'(x,n)
Construct a reduction formula, start from n=2, try to solve T'(x,n) for all integers n. (tedious like hell)
You can combine x^x with x^x^x by using the multiplication of exponential expressions adds the exponent rule where (x^x)*(x^x^x)=x^(x+x^x).
You could also write it in a more "finished" (?) form as:
d(x↑↑3)/dx = ln(x↑↑6) + ln(x↑↑4)ln(x↑↑3) + (x↑↑3)(x↑↑2)/x
Left-upper index is usually used for bottom-order tetration, for right-order hyperoperators arrow notation is basic. Left hyperoperators are rarely used though so it's specified explicitly in that case.
I have never heard that
If we have the equation y = x^^n where n is a constant, we can derive a general formula for the derivative using logarithmic differentiation:
ln y = ln x^^n
x^^n is equal to x^x^x... with n x's. This can also be rewritten as x^(x^^(n-1)). Substituting this in we get:
ln y = ln x^(x^^(n-1))
Using the rules of logs, we can move the power to the outside:
ln y = x^^(n-1) ln x
Let's try differentiating it as is and ignoring the new tetration; maybe we can find a recursive rule by doing this. For the sake of brevity, z = x^^(n-1):
ln y = z ln x
1/y dy = ln x dz + z/x dx
dy/dx = y(ln x dz/dx + z/x)
dy/dx = x^^n (ln x dz/dx + (x^^(n-1))/x)
Now we have the derivative in terms of dz/dx, the derivative for the previous value of n. We can use a base case to make this a recursive function. The easy option would be n=1, because when n=1 the function is simply y=x, and the derivative is 1. In order to make the notation simpler, let's define f like this:
f[n](x) = x^^n
(Picture [n] to mean n in the subscript, obviously I can't type that easily on a youtube comment)
Now we can write a recursive definition of f[n]'(x):
f[1]'(x) = 1
f[n]'(x) = f[n](x) (ln x f[n-1]'(x) + f[n-1](x)/x)
Let's expand the second part:
f[n]'(x) = f[n](x) ln x f[n-1]'(x) + f[n](x) f[n-1](x)/x
Now let's expand f[n-1]'(x) with the previous iteration of the function, to see if we can find a pattern to make into an explicit rule:
f[n]'(x) = f[n](x) ln x (f[n-1](x) ln x f[n-2]'(x) + f[n-1](x) f[n-2](x)/x) + f[n](x) f[n-1](x)/x
f[n]'(x) = f[n](x) ln x f[n-1](x) ln x f[n-2]'(x) + f[n](x) ln x f[n-1](x) f[n-2](x)/x + f[n](x) f[n-1](x)/x
f[n]'(x) = f[n](x) f[n-1](x) (ln x)^2 f[n-2]'(x) + ln x f[n](x) f[n-1](x) f[n-2](x)/x + f[n](x) f[n-1](x)/x
Now let's expand it one more time:
f[n]'(x) = f[n](x) f[n-1](x) (ln x)^2 (f[n-2](x) ln x f[n-3]'(x) + f[n-2](x) f[n-3](x)/x) + ln x f[n](x) f[n-1](x) f[n-2](x)/x + f[n](x) f[n-1](x)/x
Now we can see what would happen if we expand this again:
1. The first term gets another ln x and the next f function (n, n-1, n-2, ...)
2. A new term with one higher degree of ln x than the last new term and the next f function is added
Now we can create an explicit rule for this derivative:
f[n]'(x) = f[n](x) f[n-1](x) ... f[2](x) (ln x)^(n-1) + (f[n](x) f[n-1](x) + ln x f[n](x) f[n-1](x) f[n-2](x) + ... + (ln x)^(n-2) (f[n](x) f[n-1](x) f[n-2](x) ... f[1](x)))/x
f[n]'(x) = Π[k=2→n](f[k](x)) (ln x)^(n-1) + Σ[k=1→n-1]((ln x)^(k - 1) Π[h=0→k](f[n-h](x)))/x
(for all n ≥ 2)
This is a very long rule, however, it is practical to use. Let's try using it for n=3 to see if we get the same result as in the video:
f[3]'(x) = Π[k=2→3](f[k](x)) (ln x)^(2) + Σ[k=1→2]((ln x)^(k - 1) Π[h=0→k](f[3-h](x)))/x
f[3]'(x) = f[2](x) f[3](x) (ln x)^2 + ((ln x)^0 f[3](x) f[2](x) + (ln x)^1 f[3](x) f[2](x) f[1](x))/x
f[3]'(x) = f[2](x) f[3](x) ((ln x)^2 + 1/x + ln x f[1](x)/x)
f[3]'(x) = x^^2 x^^3 ((ln x)^2 + 1/x + ln x x^^1/x)
f[3]'(x) = x^^2 x^^3 ((ln x)^2 + 1/x + x ln x/x)
f[3]'(x) = x^^2 x^^3 (1/x + ln x + (ln x)^2)
Indeed, this is exactly the same as the result in the video, which means that this equation is most likely correct.
Here is the equation written in LaTeX for anyone who wants it:
www.codecogs.com/eqnedit.php?latex=\frac{d}{dx}(^nx)&space;=&space;\left\{&space;\begin{array}{ll}&space;(\ln&space;x)^{n-1}\displaystyle&space;\prod_{k=2}^{n}(^kx)&space;+&space;\frac{1}{x}\displaystyle&space;\sum_{k=1}^{n-1}((\ln&space;x)^{k-1}\displaystyle&space;\prod_{h=0}^{k}&space;(^{n-h}x))&space;&&space;\quad&space;n&space;>&space;1&space;\\&space;1&space;&&space;\quad&space;n&space;=&space;1&space;\end{array}&space;
ight%2E
“x to the superpower of y” , what if y is not an integer? For example 2^^1,5 ; how we calculate that?
@@ozanbayrak562 Using super Logarithm (inverse of Tetration)
By definition sLog2 (2^^3) = 3
NOTE: "sLog" is a notation for super Logarithm. Like how Logarithm cancels the base leaving the exponent ex. Log2 (2^3) = 3 super Logarithm does the same with Tetration leaving the super power.
We can use super Logarithm to solve non integer super powers since super Logarithm is repeated Logarithm by definition.
Let's let sLog2 (16) = 3+x
Where 0 ≤ x < 1 (represents a 0 or decimal)
sLog2 (2^^3) = sLog2 (2^2^2) => Log2(2^2^2) = 2^2
=> Log2(2^2) = 2
=>Log2(2) = 1
At this point we've taken three logs representing our integer part of the solution (given by the fact that the answer is equal to 1). We just take log again for the decimal x (the remainder of 2's that we need.)
Log2 (1) = 0
Thus sLog2 (16) = 3+0 = 3
Well let's look at what happens when we go backwards through the same process to see what happens to the remainder.
Log2 (Log2 (Log2 (Log2 (16)))) = 0
Log2 (Log2 (Log2 (16))) = 2^0
Log2 (Log2 (16)) = 2^2^0
Log2 (16) = 2^2^2^0
16 = 2^2^2^2^0 = 2^2^2 = 2^^(3+0)
The remainder adds an extra '2' to the top of the power tower and the additional 2 is raised to the power of the remainder
For 0 ≤ x ≤ 1
By definition sLog a(a^^3+x) => a^a^a^a^x
By definition of Tetration a^^3+x = a^a^^2+x = a^a^a^^1+x = a^a^a^a^^x
a^a^a^a^^x = a^a^a^a^x
a^a^a^^x = a^a^a^x
a^a^^x = a^a^x
a^^x = a^x by definition for 0 ≤ x ≤ 1
so 2^^1.5 = 2^√2 ≈ 2.6651441427
@@ryanman0083 👍
#yay Chen lu!
"this 3 is meant to be at the top-left corner"
it is at the top-left corner
Very interesting. I have following question: how can I calculate the first derivative of the function "a tetrated x" (a is positive)?
We still have no clear way of defining how to even calculate non-integer raised tetration, so good luck with that! I mean what would 2^^1.5 even mean?! I've looked into this quite a bit and there seems to be no conclusive answer.
@@DoctorT144 Using super Logarithm (inverse of Tetration)
By definition sLog2 (2^^3) = 3
NOTE: "sLog" is a notation for super Logarithm. Like how Logarithm cancels the base leaving the exponent ex. Log2 (2^3) = 3 super Logarithm does the same with Tetration leaving the super power.
We can use super Logarithm to solve non integer super powers since super Logarithm is repeated Logarithm by definition.
Let's let sLog2 (16) = 3+x
Where 0 ≤ x < 1 (represents a 0 or decimal)
sLog2 (2^^3) = sLog2 (2^2^2) => Log2(2^2^2) = 2^2
=> Log2(2^2) = 2
=>Log2(2) = 1
At this point we've taken three logs representing our integer part of the solution (given by the fact that the answer is equal to 1). We just take log again for the decimal x (the remainder of 2's that we need.)
Log2 (1) = 0
Thus sLog2 (16) = 3+0 = 3
Well let's look at what happens when we go backwards through the same process to see what happens to the remainder.
Log2 (Log2 (Log2 (Log2 (16)))) = 0
Log2 (Log2 (Log2 (16))) = 2^0
Log2 (Log2 (16)) = 2^2^0
Log2 (16) = 2^2^2^0
16 = 2^2^2^2^0 = 2^2^2 = 2^^(3+0)
The remainder adds an extra '2' to the top of the power tower and the additional 2 is raised to the power of the remainder
For 0 ≤ x ≤ 1
By definition sLog a(a^^3+x) => a^a^a^a^x
By definition of Tetration a^^3+x = a^a^^2+x = a^a^a^^1+x = a^a^a^a^^x
a^a^a^a^^x = a^a^a^a^x
a^a^a^^x = a^a^a^x
a^a^^x = a^a^x
a^^x = a^x by definition for 0 ≤ x ≤ 1
2^^1.5 = 2^2^0.5 = 2^√2 ≈ 2.6651441427...
Very clear explanation... every time i learn something 😀
Glad to hear that!
Thanks!
Loved it!!!
New notations.
Thanks Steve.. 😊😊😊
Well, I’m glad he at least admits that he makes clickbait thumbnails. Didn’t know about that notation, but already knew about tetration. Great vid
I did get the generalisation for the tetration of x to any natural index(excluding 1).Well,it goes like this,
Let x(n) be the tetration of x to a natural number n
* denotes the product of the tetrations(pi-product function)[*(k=r to k=n)
x(k)=x(r)x(r+1)x(r+2)........x(n)
# denotes summation(sigma function) [#(r=1 to r=n) x(r)=x(1)+x(2)+.........x(n)]
Then,
d[x(n)]/d(x)=(1/x)[*(k=1 to k=n)x(k)(ln(x))^n-1 + #(r=1 to r=n-1)[*(k=r to k=n)x(k)[(ln(x))]^n-r-1]]
Where #(r=1 to r=n-1)*(k=1 to k=n) denotes the summation of the products.
Put it on paper for easy understanding.Do note it is only valid for natural number greater than or equal to 2.
Just put a reply if I had gone wrong somewhere.
For the first deriv of nth tetration of x, namely
{^{n}x}I get:
(1/x)*
\sum_{j=0}^{n-1}(lnx)^{n-j-1}*
\prod_{i=j}^{n} {^{i}x}
Which can be proved by induction from the recurrence relation. Also we need to complete the pattern for the first and zeroth tetration of x
Also forgot to add x does not equal 0 to avoid divide by zero, and x not equal one, to avoid ln(1)^0 where 0^0 is also undefined
i can't believe that my friend who have never seen tetration or heard about thought of this concept and wrote it in the same annotation of this and chose to call it superpostion .then he sarted studyng it as a functon and he got some pretty cool stuff .bt he was stuck on a problem.while searching the net for a solution he fond the same concept in the name of tetration and it shook us how similar his invention is to it .
Do a vid about integrating e^((x^x)*ln(x)) next please
you can't integrate x^x, so pretty sure you cant integrate x^(x^x)(which is what you asked here)
dolev goaz That's no fun.
@dolev goaz:
Since we're talking mathematics here, I sincerely wonder: *"do you have proof for your two claims?"*
(The claims I mean: 1: "you can't integrate x^x" and 2: "so you can't integrate x^(x^x)" )
For x>0 both functions (i.e. exp(ln(x)*exp(ln(x))) and exp(ln(x)*exp(ln(x)*exp(ln(x)))) ) are continuous and differentiable, so why wouldn't you be able to integrate them?
More precisely, their indefinite integrals aren't expressible in terms of elementary functions.
YES! He can't tease us like that!!
I think I have the general formula where I set x↑↑0=1 (Is that right?).
First a recursion formula:
d/dx(x↑↑n) = x↑↑n · (d/dx(x↑↑(n-1)) · lnx + x↑↑(n-1) · 1/x)
After applying that a few times I found a pattern, so here's the formula (I think):
d/dx(x↑↑n) = 1/x · sum_(k=0 to n-1) of [(lnx)^(n-1-k) · product_(m=k to n) of (x↑↑m)]
d/dx(x↑↑0) = d/dx(1) = 1/x · sum_(k=0 to -1) of [(lnx)^(-1-k) · product_(m=k to 0) of (x↑↑m)]
= 1/x · 0
= 0
d/dx(x↑↑1) = d/dx(x) = 1/x · sum_(k=0 to 0) of [(lnx)^(-k) · product_(m=k to 1) of (x↑↑m)]
= 1/x · [(lnx)^(0) · product_(m=0 to 1) of (x↑↑m)]
= 1/x · x↑↑0 · x↑↑1 = 1/x · 1 · x
= 1
d/dx(x↑↑2) = d/dx(x^x) = 1/x · sum_(k=0 to 1) of [(lnx)^(1-k) · product_(m=k to 2) of (x↑↑m)]
= 1/x · [(lnx)^(1) · product_(m=0 to 2) of (x↑↑m) + (lnx)^(0) · product_(m=1 to 2) of (x↑↑m)]
= 1/x · [lnx · x↑↑0· x↑↑1 · x↑↑2 + x↑↑1 · x↑↑2]
= 1/x · [lnx · 1· x· x^x + x · x^x]
= lnx· x^x + x^x
= x^x · (lnx+1)
d/dx(x↑↑3) = d/dx(x^x^x) = 1/x · sum_(k=0 to 2) of [(lnx)^(2-k) · product_(m=k to 3) of (x↑↑m)]
= 1/x · [(lnx)^(2) · product_(m=0 to 3) of (x↑↑m) + (lnx)^(1) · product_(m=1 to 3) of (x↑↑m) + (lnx)^(0) · product_(m=2 to 3) of (x↑↑m)]
= 1/x · [ln²x · x↑↑0 · x↑↑1 · x↑↑2 · x↑↑3 + lnx · x↑↑1 · x↑↑2 · x↑↑3 + x↑↑2 · x↑↑3]
= 1/x · [ln²x · 1 · x · x^x · x^x^x + lnx · x · x^x · x^x^x + x^x · x^x^x]
= ln²x · x^x · x^x^x + lnx · x^x · x^x^x + x^x · x^x^x · 1/x
= x^x^x · x^x · (ln²x+lnx+1/x)
Instead of plugging in exp(xln(x)) each and every time to replace x^x,
why not just use the power rule for differentiating the function power of another function,
thus if y = u^v, then
dy = vu^(v-1)du + (u^v)ln(u)dv
So if y = x^x, then
dy = xx^(x-1)dx = x^xln(x)dx
= x^xdx = x^xln(x)dx
= x^x[1 + ln(x)]dx
Do a video on solving tetration equations please as well as tetration with imaginary numbers and taking the super-root of those numbers.
Tetration! Knuth Arrow Notation! #YAY
Note: We can use a reduction formula for the derivative of x^^n
d/dx (x^^n) = d/dx (e^((x^^(n-1)ln(x)))) = (x^^n)((x^^(n-1))/x+ln(x) d/dx (x^^(n-1)))
Or saying
D(n) = (x^^n)((x^^(n-1))/x+ln(x)*D(n-1)))
I think the proper way of pronouncing tetration would be „x tetrated to n“ or „x tetrated to the n-th power“. And btw, just as we write x^n for „x to the power of n“, i think we also could simply write x^^n for „x tetrated to n“.
i like saying "n tet x" cause I'm lazy
I think "x tetrated by n" or "n tetrating x" both have a nice sound to them
Video starts at 3:36 for those who already know what tetration and arrow notation are.
This is what I got, though I’m not sure if correct! It’s technically an explicit function 😂
d(x || n)/dx = E(k=1 to n) [ {((ln(x))^(n-k))/x} * P(i=k-1 to n) [x || i] ]
Notation
E ~ the sigma summation function
P ~ the sigma pin function
|| ~ the double arrow function
Domain:
For n within positive integers and assuming that x || 0 = 1
I'm from Indonesia and i find this interesting!
The rule is x↑↑n = (x↑↑n) * (d/dx ( x↑↑(n-1) * ln(x) + x↑↑(n-1) *(1/x))
Thats the best you can do
its basically the original function multiplied by the product rule of x↑↑(n-1) and ln(x)
Change the 3's at graham number to x and find the derivative of it
muhammad fatih
omg, brutal
I don’t think that it’s really possible, it would be way too long and there would be mistakes
@@titouanvasnier2347 yea i know, im just messin around, and chen lu think he can take everything on so i challenge hin
The derivative of Graham's number is definitely 0. It's a constant.
dlevi67 got’em
La solución a lo del final sería algo así como:
d/dx(x↑↑n)=fact(a=1 to n-1, x↑↑a)*sum(b=0 to n-1,(ln(x))^b)
No es tan simple como la otra, pero creo que sirve, no?
En fin, no lo he comprobado.
The solution to the final stuff would be something like:
d/dx(x↑↑n)=fact(a=1 to n-1, x↑↑a)*sum(b=0 to n-1,(ln(x))^b)
It's not so simple as the other one, but I think it works, didn't it?
Well, I haven't confirmed it.
You forgot the term 1/x ;-)
It can further simplify by bringing the 2 from the natural log of x in front of it then add it to the other one
Tetration in general needs a complete follow up video
Wow. I learned a new thing today. Thanks.
you cannot write x = exp(ln(x)), without defining the values x can take since the ln function is only defined on ]0 ; +infinity[ !!
@Gerben van Straaten if i'm not mistaken this function is also defined for negative rational numbers that can be written as -a/b where b is an odd integer, however the fact that it would not be continuous still stands
As always he made it waaaayy longer than it needed to be!
To answer the last question in the video, just from generalization of the result, I’m guessing:
d/dx (x^^n) = x^^(n-1).x^^n.(1/x + ln(x) + ln(x)^(n-1))
No. This is not right.
Brother u are really a genius !!!💖👏👏👌👑
I was also thinking of using substitution and logarithmic differentiation. So solve for x^u, where u= x^x. Use logarithmic differentiation for u first, then after finding u’, use it to find x^u. I think this is long and messy though.
Now Dr. Peyam has to make a video about the generalization d/dx(x↑↑a), where a represents any rational number!
One thing my Maths Sir once told about differentiating x^x is that "first differentiate it as constant^variable then differentiate it as variable^constant and add them that is the derivative of x^x" idk if it works for any tetration
Funny is that there is a rule that alows us to derivate anything no matter how tetrated or stacked it is. Basically instead of doing what he did you can derivate the whole thing a function at a time from outside to inside and write the product of those partial results
I’ve taken 3 levels of calc and that notation never came up. Great video
Solve for this; Cow tied at the corner of a circular field is able to graze 3/4th of the field. What is ratio of length of its rope and radius of the field?
I got the answer... yeah, I watched the video quite late, but I found the answer all on my own, and then watched the rest of the video, I also got the answer to that last question you asked 😄( it might not seem like a big deal, but I am in 10th grade 😅😅😅 ) it was fun doing it... keep posting such videos 👍🏻👍🏻👍🏻
ok so i just did probably some of the most ugly math ever and found a general form for the derivative of x^^n. so ill try to write it out here for anyone who cares:
I'll start out with just the explicit equation, and then explain the notation.
d/dx [x^^n] = (1/x * Σ [ t!(n,k) * (ln(x))^(n-1-k) ] ) + t!(n,2)*( (ln(x))^(n-1) + (ln(x))^(n-2) ), where Σ is from k=2 to k=n-1
first of all, x^^n means x tetrated to the n, or ⁿx, hopefully that was clear
ok so, the t!(x,y) is a notation i made up that stands for "tetration factorial", where it multiplies many x^^(something), where (x,y) is the range of the (something) in question
for example: t!(4,1) = (x^^4)*(x^^3)*(x^^2)*(x^^1)
or: t!(5,3) = (x^^5)*(x^^4)*(x^^3)
technically, ive only tested this for the derivatives of x^^5, x^^6, and x^^7, but yeah. do with this information what you will. I know its probably the most hideous way of expressing this but whatever.
edit: heres a desmos graph that proves that it works for x^^7 if youd like to see www.desmos.com/calculator/ufqbmbhbzs
Wow, that's really cool! Now I'm curious if it's possible to differentiate n↑↑x.
We would first have to figure out what x^^1.5 would even mean. I've done a lot of research on this and there seems to be no conclusive answer.
Since multiplying a number by a non-integer number and taking a number to a non-integer power are both possible, is tetrating a number to an non-integer number possible? Is it also possible for the other infinitely many arithmetic operations?
y = x^f(x) -> log(y) = f(x) log(x) -> y'/y = f'(x) log(x) + f(x)/x ->
y' = y (f'(x) log(x) + f(x)/x)
this is the derivation of the recursive formula below
Tetration is the first step to big numbers... The goes pentation, then hexation, then Knut's notation, then Conway's notation, and then the fast-growing hierarchy, which will bring you very far away from what is considered 'normal'.
Please make videos on mulrivariable calculus
Man it would be cool to have this guy as a calculus teachers
Do the derivity of x⬆️⬆️x
I was unable to come up with an explicit formula for (d(x↑↑n)/dx), but I was able to come up with a recursive formula (I hope). If someone could tell me if the formula is correct or how to make it explicit, that would be fantastic.
Formula:
d(x↑↑n)/dx = (x↑↑n)(ln(x)(d(x↑↑(n-1))/dx)+((x↑↑n)/x))
Not only he uses the up arrow, he rewrote the answer in 3 different ways to make the video 10+ mins. My man is taking the wise words from Pewdiepie.
thinking about fractional tetration
really enjoy your style 👍
I would like to know if tetration has properties. like, if it was a x^2 * x^3 right there, you would be able to add the exponents, but it dosn't seem to be the case with tetration
Thanks, I've always wondered this!
I am ... amazed
You say chain rule, I hear Chen Lu.😂
Me too I heard chen lu
Awesome! Thank you!!
Awesome as always!
I found the formula, bear with me
d/dx x||n (I cant draw arrows lol) is equal to
[Π (i goes from 2 to n) x||i] * [(Σ (a goes from 0 to n-1) (lnx)^a ÷ Π (q goes from 0 to n-a-2) x||q ]
Yeah it's kinda lengthy but try to write it down on a paper and check, it works
I found a quite complicated formula for general integration of tetration functions in wikipedia, here: en.wikipedia.org/wiki/List_of_integrals_of_exponential_functions
It is given as exactly 20th example from the top; the formula involves incomplete gamma function and some parameters (n, m, i, j), but it's still not quite clear to me. I have just subscribed your channel :-) Could you devote some time to this formula in one of your future lectures? I mean, how it has been found and how it can be used In practice? Anything that can make it more familiar and clear?
It's cool and all, but the page contains integrals. Can you find one which lists the formula for derivatives?
@@nishilsheth9076 You know, derivatives can be calculated using quite simple set of a few rules. They include: a formula for derivative of the product (1), derivative of the quotient (2), derivative of a composite function (3), derivative of a power (4), and derivative of an inverse function (5). So here you are:
(1) [f(x)*g(x)]' = f'(x)*g(x) + g'(x)*f(x), or - more simply, (fg)' = f'g + g'f
(2) [f(x)/g(x)]' = [f'(x)*g(x) - g'(x)*f(x)] / [g(x)]^2, more simply (f/g)' = (f'g - g'f)/g^2
(3) when u(x) = f(g(x)), u'(x) = g'(x) * f'(u), for example [ln(sinx)]' = (1/sinx)*(-cosx) etc.
(4) [f(x)^g(x)]' = [f(x)^g(x)] * {[g'(x) * ln f(x)] + [g(x)/f(x) * f'(x)]}, or more simply:
(f^g)' = (f^g) * (g'* lnf + g*f'/f ) - which is an applied formula for derivative of a composite function after making an identity transformation: f^g = e^(g*lnf)
(5) when y=f(x) and x=g(y), g'(y)=1/f'(x).
For example when x=y^y and y=ssrt(x), (y^y)* (1+lny) = 1/[ssrt(x)]', which yields [ssrt(x)]' = 1/[y^y * (1+ lny)], so:
when f(x) = ssrt(x), f'(x) = 1/ [x + ln(ssrt(x)], since when y= ssrt(x), y^y = x by definition
When using these five formulas, we can determine a derivative of any function expressed with a finite formula.
I'll try it. So x^(x^x) = y, e^(x^x)ln(x) = y, e^e^(ln(ln(x))+xln(x))=y
Derivative is x^(x^x) * (x^x)ln(x) * (1/xln(x) + ln(x) + 1)
Which can be simplified to x^(x^x +x) *(1/x + ln(x)^2 + ln(x))
Standard derivation of e^ax = a.e^ax, because a is constant. But at video - 5:00, X^x is variable.
This is not clear. I never attempted.
d(e↑x)/dx = e↑x; one of the reasons e is a special number for calculus
x = e↑(ln x); an identity that synergizes well with the previous rule
d(f(g(x)))/dx = f'(g(x)g'(x); the chain rule of calculus.
x↑x = e↑(x ln x)
Finding the derivative of x ln x requires the product rule: d[f(x)g(x)]/dx = f'(x)g(x) + f(x)g'(x)
1* ln x + x * 1/x
ln x + 1
d(x↑x)/dx = (1 + ln x) e↑(x ln x)
which simplifies to (1 + ln x) x↑x
So what's the general form of d/dx(x↑↑n )?
This already isn't true for x↑↑2.
Man, I've been studying this relation for hours, and something seems to pop up.
First, I've found a reduction formula.
Second, I've seen this behavior pop up:
The derivative of x^x is x^x(1+ln(x)), normal, right?
x^x^x is x^x^x*x^x(1/x+ln(x)+ln^2(x)), the last two are normal.
x^x^x^x is x^x^x^x*x^x^x*x^x(1/(x^x*x)+ln(x)/x+ln^2(x)+ln^3(x)), we see the behavior 1 over x^x*x, x, 1, 1
x^^5 is with behaviour x^x^x*x^x*x, x^x*x, x, 1, 1
So it's for now the best to study why this behavior happens and then we'll have a formula to calculate d/dx (x^^n)
I think the behavior happens because when we apply my reduction formula, we have new tetrations of x which add on to the 1/x, and the previous terms are preserved.
D(3)=(x^^3)((x^^2)/x+ln(x)*x^x(1+ln(x)))
= x^^3*x^^2(1/x+ln(x)+ln^2(x))
The cancellation adds 1/x
D(4)=(x^^4)((x^^3)/x+ln(x)*x^^3*x^^2(1/x+ln(x)+ln^2(x)))
= x^^4*x^^3(1/x+ln(x)^x^^2(1/x+ln(x)+ln^2(x)))
But then we have to divide x^^2 which adds to the 1/x term, so it becomes x^x*x
= x^^4*x^^3*x^^2(1/x^x*x+ln(x)(1/x+ln(x)+ln^2(x)))
And the other terms are preserved.
So you see what I mean?
Why is my comment not showing up?!?
Anyway, in the comment, I wrote this:
"We can use a reduction formula for the derivative of x^^n
d/dx (x^^n) = d/dx (e^((x^^(n-1)ln(x)))) = (x^^n)((x^^(n-1))/x+ln(x) d/dx (x^^(n-1)))
Or saying
D(n) = (x^^n)((x^^(n-1))/x+ln(x)*D(n-1)))"
x↑↑2*x↑↑3*x↑↑4*...*x↑↑n*(1/x+ln(x)+ln(x)^2+ln(x)^3+....+ln(x)^(n-1))
I am here to comment with a gaming account again lol
The derivative of the tetration of x is equal to the tetration of x times the quantity ln x +1.
I didn't even know that I could write X^(X^(X)) this way lol