A number theory problem by Dev Sharma

Which is greater?

300 ! or 10 0 300 ? 300! \text{ or } 100^{300} ?

300 ! 300! 10 0 300 100^{300} Both are equal

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

13 solutions

Otto Bretscher
Aug 21, 2015

The inequality n ! > ( n / 3 ) n n!>(n/3)^n holds for all positive integers n n ; apply this to n = 300 n=300 .

Note: One way to prove the inequality is: ln ( n ! ) = k = 2 n ln ( k ) 1 n ln ( x ) d x > n ( ln ( n e ) ) > n ( ln ( n 3 ) ) = ln ( ( n 3 ) n ) \ln(n!)=\sum_{k=2}^n \ln(k)\geq\int_{1}^{n}\ln(x) dx >n(\ln(\frac{n}{e} ))>n(\ln(\frac{n}{3}))=\ln((\frac{n}{3} )^n)

Another way to prove the inequality is:
We can give a proof by induction on n n , assuming that n ! > n n 3 n n!>\frac{n^n}{3^n} . Now ( n + 1 ) ! > ( n + 1 ) n n 3 n = ( n n + 1 ) n ( n + 1 ) n + 1 3 n = 1 ( 1 + 1 / n ) n ( n + 1 ) n + 1 3 n > ( n + 1 ) n + 1 3 n + 1 (n+1)!>(n+1)\frac{n^n}{3^n}=(\frac{n}{n+1})^n\frac{(n+1)^{n+1}}{3^n}=\frac{1}{(1+1/n)^n}\frac{(n+1)^{n+1}}{3^n}>\frac{(n+1)^{n+1}}{3^{n+1}} since ( 1 + 1 n ) n < e < 3 (1+\frac{1}{n})^n<e<3 . We have in fact shown that n ! > n n e n n!>\frac{n^n}{e^n} .

Oh, that is really nice! Thanks for sharing a calculus area approximation solution!

Calvin Lin Staff - 5 years, 9 months ago

Oh, Wow! That was just awesome!

Kishore S. Shenoy - 5 years, 8 months ago

Great solution

Sachin Sharma - 4 years, 9 months ago

In fact you can prove (I don't know how, though!) that ( n 3 ) n n ! ( n 2 ) n \left(\dfrac n 3\right)^n\le n!\le\left(\dfrac n 2\right)^n

Can you attempt a proof? Thanks in advance!

Kishore S. Shenoy - 4 years, 9 months ago

can you please explain from where the inequality comes from in the equation? I'm a beginner so, please don't mind if this is a silly question.

Subham Agarwal - 2 years, 11 months ago

Log in to reply

I was also struggling with the intermediate steps of the proof. This is what I came up with:

  • First equality: Distributing the logarithm onto the factors of the factorial (converting the multiplications to sums) and omitting ln ( 1 ) = 0 \ln(1) = 0 from the result.
  • First inequality: The left side is the right Riemann sum (approximating the integral) for an increasing function, yielding an overestimate. Hence the greater than or equal sign.
  • Second inequality: Maybe someone else has a shorter proof. I was only able to show this inequality using quite a few intermediate steps: First replace the x x inside the logarithm by ( x e e ) (\frac{x}{e} \cdot e) , then split the logarithm of the product into a sum of logarithms, integrate both parts (using the integral formula for the ln \ln function) and you'll end up with the integral being equal to n ln ( n e ) n e + n + 1 n \cdot \ln(\frac{n}{e}) - \frac{n}{e} + n + 1 . Given that n e + n + 1 -\frac{n}{e} + n + 1 is always greater than zero for any positive integer n n , this yields the inequality.
  • The final two steps are straightforward.

Andreas Keil - 2 years, 11 months ago

Log in to reply

Indeed, the first inequality follows directly from "right Riemann sum of an increasing function", and its a nice way to approximate the discretized sum of an integrable function.

For the second inequality, from integration by parts applied directly, 1 n ln x d x = [ x ln x x ] 1 n = n ( ln n ln e ) 1 ( ln 1 1 ) = n ln n e + 1 > n ln n e . \int_1^n \ln x \, dx = [x \ln x - x]_1^n = n ( \ln n - \ln e ) - 1 (\ln 1 -1 ) = n \ln \frac{n}{e} + 1 > n \ln \frac{n}{e}. I'm not quite sure where your extra terms came from.

Calvin Lin Staff - 2 years, 11 months ago

I'm sorry for the long delay; I'm not spending much time on Brilliant these days. We can give a proof by induction on n n , assuming that n ! > n n 3 n n!>\frac{n^n}{3^n} . Now ( n + 1 ) ! > ( n + 1 ) n n 3 n = ( n n + 1 ) n ( n + 1 ) n + 1 3 n = 1 ( 1 + 1 / n ) n ( n + 1 ) n + 1 3 n > ( n + 1 ) n + 1 3 n + 1 (n+1)!>(n+1)\frac{n^n}{3^n}=(\frac{n}{n+1})^n\frac{(n+1)^{n+1}}{3^n}=\frac{1}{(1+1/n)^n}\frac{(n+1)^{n+1}}{3^n}>\frac{(n+1)^{n+1}}{3^{n+1}} since ( 1 + 1 n ) n < e < 3 (1+\frac{1}{n})^n<e<3 . We have in fact shown that n ! > n n e n n!>\frac{n^n}{e^n} . I hope this brief explanation is helpful...

Otto Bretscher - 2 years, 8 months ago
Antony Moyalan
Aug 22, 2015

okay so what i pretty much did was just picture the expansion in my mind. 100^300 is 300 100s right? and 300! is every number counting down from 300. so in this theres 200 numbers greater than 100 and only 100 numbers less than 100 . so from logical deduction i assume you can figure out that 300! is larger

Approach is incorrect, kind of. If I ask you 215! And 100^215, your answer might have been wrong

Aman Verma - 4 years, 9 months ago

Yeah, same way here.

Drex Beckman - 5 years, 5 months ago

Same here.

Sattik Biswas - 5 years, 2 months ago

how would you do it for 100^200 and 200! ??

Shivam Singh - 4 years, 11 months ago

I agree, I used the same technique to guess. Because 100 multiplied by all the integers from 101 to 300 must be greater than 100^200. I estimated that the remaining integers under 100 would give at least another 50 multiples of 100 or more (e.g. 2 99 > 100, 3 98 >100, etc.). Then I confidently guessed that I have made enough multiplications greater than 100 to make up for the last 50 multiples. Still, it is an educated guess.

In response to another comment, using the same technique for 100^200 and 200! I probably would have determined that while 200! > 100^150, I don't think I would have the confidence to say that 200! > 100^200, so a more thorough check would have been needed. :)

Lilli Hooper - 4 years, 1 month ago

What I thought that 300! contains 9,29, ...…99,199 and these get multipied. So there product will be larger than 300 100s( 100^300) so that was mine logic

Riya Verma - 3 years, 4 months ago
Curtis Clement
Aug 14, 2015

We can use Stirling's formula to solve the problem quite nicely. For those of you that aren't familiar with the formula I will state it below: lim n n ! 2 π n . ( n e ) n \lim_{n \rightarrow\infty} n! \rightarrow\sqrt{2 \pi n}. ( \frac{n}{e}\ )^n With (geometric) convergence from above (i.e. n ! > R H S \ n! > RHS for n > 0 \ n > 0 ) 300 ! > 600 π . ( 300 e ) 300 > 43 × 10 0 300 > 10 0 300 \Rightarrow\ 300! > \sqrt{600 \pi} . (\frac{300}{e})^{300} > 43 \times\ 100^{300 } > 100^{300} 300 ! > 10 0 300 \therefore\ 300! > 100^{300}

I don't see the 43 ????

Eu Mim - 5 years, 3 months ago

Log in to reply

600 π = 43.416 43 \sqrt{600\pi} = 43.416 \approx 43 .

Adam Blakey - 5 years, 3 months ago
Pranesh Hari
Nov 17, 2016

I just used a scientific calculator to know the answer. Here is what I did: let x be the geometric mean of first 300 natural numbers, then we can alternatively write 300! = x 300 x^{300} so if x > 100 then 300! > 10 0 300 100^{300} and vice versa, so the geometric mean x = 300 ! 300 \sqrt[300]{300!} = 111.758 (appox) so x > 100 therefore, 300! > 10 0 300 100^{300}

Peter De
Aug 25, 2018

Simplistically:-

1.) log(100^300)= 300 x log(100)=300 x 2=600

2.) log(300!)=log(300)+log(299)+....+log(100)+....+log(10)+.....+log(1)

Log to the base 10 of 100 is 2, of 10 is 1, and of 1 is zero

now pairing: log(300)+log(1) >2
log(299)+log(2)> 2 ...... log(201)+log(100)>2 ..... log(151)+log(150)>2

Therefore all pairs are >2 Therefore log(300!) > 2 x 300 >600

Therefore 300! > 100^300 q.e.d.

Prokash Shakkhar
Dec 15, 2016

The question is to find out the larger number. Okay, If we plug ? ? as relational operator between 300 ! 300! and 10 0 300 100^{300} then the desired form is,

(300!) {\color\red{?}} (100^{300}) \Rightarrow {log_{100} (300!)} {\color\red{?}} {300log_{100} (100)} \Rightarrow{ \sum_{x=300}^1 log_{100} x} {\color\red{?}} {300} \Rightarrow { log_{100} (300) + log_{100} (299) + log_{100} (298) \ldots + log_{100} (1) } {\color\red{?}} (300 ) Now, look carefully that, the number of terms in left side is exact to the number 300 in right side.. So, left side = right side will be possible if and only if every term of l o g 100 x log_{100} x is 1 1 .. But it is not possible. Because l o g 100 ( 300 ) log_{100} (300) , l o g 100 ( 299 ) log_{100} (299) ....... upto l o g 100 ( > 100 ) log_{100} (>100) , every term will be > 1 >1 .. So, x = 300 1 l o g 100 x \sum_{x=300}^1 log_{100} x > 300. \Rightarrow \boxed{\color\red{300! > 100^{300}}}

"So, left side = right side will be possible if and only if every term is 1". False. Because ... every term will be >1. And for every x<100 log_100x<1. And then?

Antonio Rinaldi - 4 years, 4 months ago
Ryan Ta
Oct 15, 2016

We can prove by induction that n ! > 10 0 n n! > 100^n for all integers n 269 n \ge 269 .

Base step: At n = 269 n=269 , we have that 269 ! 2.467 × 1 0 538 > 1 × 1 0 538 = 1 0 269 269! \approx 2.467 \times 10^{538} > 1 \times 10^{538} = 10^{269} .

Induction step: At n = k n=k (for some k > 269 k > 269 ), assume that k ! > 10 0 k k! > 100^k . Then we have that ( k + 1 ) ! = k ! ( k + 1 ) > 10 0 k ( k + 1 ) > 10 0 k 270 > 10 0 k 100 = 10 0 k + 1 . (k+1)! = k!(k+1) > 100^k(k+1) > 100^k \cdot 270 > 100^k \cdot 100 = 100^{k+1}. Since ( k + 1 ) ! > 10 0 k + 1 (k+1)! > 100^{k+1} , this completes the induction proof.

Disclaimer: For the base step, I admittedly experimented with trial-and-error on Wolfram Alpha to find that the smallest number n n for which n ! > 10 0 n n! > 100^n is true is 269 269 .

Mainak Samanta
Jun 4, 2018

(N/2)^2 > N! > (N/3)^3

Everyone SpamOne
Jan 1, 2018

Kamenetsky's formula floor(\(\frac{log(2𝜋x300)}{2}\) + 300(log(300)-log(e))) + 1 gives 615 digits in 300! And 300log(100) + 1 gives 601 digits in 100^300. Note that normally to find the number of digits in an exponent the floor function would be used, but in this case, taking the log of this number gives an integer, so floor function is not necessary. 300! > 100^300

Richard Levine
Apr 15, 2016

300! = 1(300!) = 100^300(.01^300)300! That's greater than 100^300 if 300! has greater than 600 digits.

Number of digits in X = CEILING[log(X)] CEILING[log(N!)] = CEILING[log(N(N-1)...1)] = CEILING[log(N)+log(N-1)+...+1] Number of digits in 300! = CEILING[log(300) +log(299) ... +log(1)]

log(300) = 2.477; log(100) = 2; log(10) = 1. If log(X) was linear, we'd have approximately 300! with 201(2.23) + 89(1.5) digits, which is 448 + 133 = 581 digits. However, log(X) is NOT linear. log(200) = 2.30; log(49) = 1.69. So we have at least 202(2.3) + 89(1.6) digits = 464 + 142 = 606 digits. That's greater than 600. Therefore, (.01^300)300! > 1 and 300! > 100^300.

Debanjan Dey
Aug 23, 2015

As per me, 300 ! = 300 × 299 × . . . × 2 × 1 300!\quad =300\quad \times\quad 299\quad \times\quad ...\quad\times\quad2\quad\times\quad 1 = 150 × 150 × . . . 300 t i m e s =\quad 150\times 150\times ...\quad 300\quad times

= 150 300 ={ 150 }^{ 300 } , obviuosly > 100 300 >{ 100}^{ 300}

Debanjan, your equality 300!=150^300 is not true. Try it for 8! and you will see that it does not equal to 4^8 (which is greater).

Guillaume Barrette - 5 years, 9 months ago

Log in to reply

Thanks! My answer was rather based on assumption. I was doubtful for actual case.

Debanjan Dey - 5 years, 9 months ago

Simple expansion in mind. Imagine that both are expanded. Consider the factorial term, and in that, consider the numbers which are above 100, divide it in such a way that those numbers doesn't go below 100 and you have at least 200 numbers between 1 and 3. Use those to multiply the numbers which are below 100. Now you have 300 numbers which are distinctively more than 100. That means, it's more than the other term.

Moderator note:

Use those to multiply the numbers which are below 100. Now you have 300 numbers which are distinctively more than 100.

This is not obvious. Can you show the relevant working please?

can u give a soln othr dan ds so dt sum1 who is nt a math student could also solve it cuz ds factorial n sterling formula r need to b remembered...........hw abt if take log of both....or any soln using general sense of humour............

Sonica Yadav - 5 years, 10 months ago
Ian Williams
Aug 21, 2015

I arrived at the solution through logical deduction, Both the solutions can be computed though a string of exactly 300 sequential multiplications. Thus the number with the highest mean factor of multiplication must be the largest.

the mean for 300! =(1+300)/2=150.5

while 100^300=100

300!>100^300

That is not true. For example, 20 ! < 9 20 20! < 9 ^ {20} , even though 10.5 > 9 10.5 > 9 .

Calvin Lin Staff - 5 years, 9 months ago

Log in to reply

Yup u are right

Riya Verma - 3 years, 4 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...