Floor problems

How many integers from 1 to 1000 inclusive are there in the form 2 n + 3 n \left\lfloor 2n \right\rfloor +\left\lfloor 3n \right\rfloor for any positive real number n n ?


The answer is 800.

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.

5 solutions

Maurice Patel
Feb 1, 2015

Given that

\hspace{10mm} 2 n + 3 n \left\lfloor 2n \right\rfloor + \left\lfloor 3n \right\rfloor

where n is any real integer.

We shall consider the following four cases:-

i) 0 n < 0.33 \hspace{3mm} 0\leq\; n \; <\; 0.33

0 2 n < 0.66 \hspace{5mm} \therefore\; 0\leq 2n <\;0.66 \hspace{4mm} and 0 3 n < 0.99 \hspace{5mm} 0\leq 3n <\;0.99 \hspace{4mm}

2 n = 2 n \hspace{5mm} \therefore\;\left\lfloor 2n \right\rfloor = 2n \hspace{4mm} and 3 n = 3 n \hspace{5mm} \left\lfloor 3n \right\rfloor = 3n

\hspace{5mm} So the number is of form:-

2 n + 3 n = 2 n + 3 n = 5 n \hspace{7mm} \left\lfloor 2n \right\rfloor + \left\lfloor 3n \right\rfloor = 2n + 3n = 5n

\hspace{7mm} i.e. 5 , 10 , , 1000 \hspace{3mm} 5,10,\dots,1000

\hspace{5mm} We can easily find no of terms in this A.P. which are equal to 200 \hspace{3mm}\boxed{200}

ii) 0.33 < n < 0.5 \hspace{3mm} 0.33<\;n < 0.5

0.66 < 2 n < 1 \hspace{5mm} \therefore\; 0.66 < 2n <1 \hspace{4mm} and 0.99 < 3 n < 1.5 \hspace{5mm} 0.99< 3n <1.5 \hspace{4mm}

2 n = 2 n \hspace{5mm} \therefore\;\left\lfloor 2n \right\rfloor = 2n \hspace{4mm} and 3 n = 3 n + 1 \hspace{5mm} \left\lfloor 3n \right\rfloor = 3n + 1

\hspace{5mm} So the number is of form:-

2 n + 3 n = 2 n + 3 n + 1 = 5 n + 1 \hspace{7mm} \left\lfloor 2n \right\rfloor + \left\lfloor 3n \right\rfloor = 2n + 3n + 1 = 5n + 1

\hspace{7mm} i.e. 1 , 6 , , 996 \hspace{3mm} 1,6,\dots,996

\hspace{5mm} This is also an A.P. which is having 200 \hspace{3mm}\boxed{200}\hspace{3mm} terms.

iii) 0.5 n < 0.66 \hspace{3mm} 0.5\leq\;n <0.66

1 2 n < 1.32 \hspace{5mm} \therefore\; 1 \leq 2n <1.32 \hspace{4mm} and 1.5 3 n < 1.98 \hspace{5mm} 1.5\leq 3n <1.98 \hspace{4mm}

2 n = 2 n + 1 \hspace{5mm} \therefore\;\left\lfloor 2n \right\rfloor = 2n + 1\hspace{4mm} and 3 n = 3 n + 1 \hspace{5mm} \left\lfloor 3n \right\rfloor = 3n + 1

\hspace{5mm} So the number is of form:-

2 n + 3 n = 2 n + 1 + 3 n + 1 = 5 n + 2 \hspace{7mm} \left\lfloor 2n \right\rfloor + \left\lfloor 3n \right\rfloor = 2n + 1 + 3n + 1 = 5n + 2

\hspace{7mm} i.e. 2 , 7 , , 997 \hspace{3mm} 2,7,\dots,997

\hspace{5mm} This A.P. is having 200 \hspace{3mm}\boxed{200}\hspace{3mm} terms.

iv) 0.66 < n < 1 \hspace{3mm} 0.66< n <1

1.32 < 2 n < 2 \hspace{5mm} \therefore\; 1.32 < 2n <2 \hspace{4mm} and 1.99 < 3 n < 3 \hspace{5mm} 1.99< 3n <3 \hspace{4mm}

2 n = 2 n + 1 \hspace{5mm} \therefore\;\left\lfloor 2n \right\rfloor = 2n + 1\hspace{4mm} and 3 n = 3 n + 2 \hspace{5mm} \left\lfloor 3n \right\rfloor = 3n + 2

\hspace{5mm} So the number is of form:-

2 n + 3 n = 2 n + 1 + 3 n + 2 = 5 n + 3 \hspace{7mm} \left\lfloor 2n \right\rfloor + \left\lfloor 3n \right\rfloor = 2n + 1 + 3n + 2 = 5n + 3

\hspace{7mm} i.e. 3 , 8 , , 998 \hspace{3mm} 3,8,\dots,998

\hspace{5mm} This A.P. is having 200 \hspace{3mm}\boxed{200}\hspace{3mm} terms.

\hspace{4mm}\Rightarrow Answer is 800 \hspace{1mm}\boxed{800}

Woah.... Kind of overrated.....

Julian Poon - 6 years, 4 months ago

Log in to reply

I have mentioned it in a comment on Luke's Solution. "less than 5 min to solve but more than 25 minutes to write solution".

PS- It was my first solution so i wasn't knowing in how much depth one has to write the Solution.

Maurice Patel - 6 years, 4 months ago

Log in to reply

Great job with your first solution! Very well written and explained :)

Calvin Lin Staff - 6 years, 4 months ago

How did you get the inequality?

Luke Zhang - 6 years, 4 months ago

Log in to reply

I am considering this inequality 0 < n < 1 \hspace{3mm}0 < n < 1

0 < 2 n < 2 \hspace{5mm} \therefore\; 0 < 2n < 2 \hspace{4mm} and 0 < 3 n < 3 \hspace{5mm} 0 < 3n < 3 \hspace{4mm}

So we can see that

m a x 2 n = 2 n + 1 \hspace{3mm}max\hspace{1mm}\left\lfloor 2n \right\rfloor = 2n + 1\hspace{4mm} and m a x 3 n = 3 n + 2 \hspace{5mm}max\hspace{1mm}\left\lfloor 3n \right\rfloor = 3n + 2

Now after knowing this it is easy to divide ( 0 , 1 ) \hspace{2mm}(0,1) \hspace{2mm} into:-

1) two parts for 2 n \hspace{2mm}2n\hspace{2mm}

2) three parts for 3 n \hspace{2mm}3n\hspace{2mm}

So we get total 4 4 partitions of ( 0 , 1 ) \hspace{2mm}(0,1) \hspace{2mm} as mentioned in the solution.

Is it clear now?

Maurice Patel - 6 years, 4 months ago

Log in to reply

nope not clear.

Luke Zhang - 6 years, 4 months ago

Log in to reply

@Luke Zhang Is it possible for you to post the full solution. Do you have other means of communication? HOw did u get the inequality for the 0.33,0.5,0.66,1. And how do u get 2n +1, 3n+1, etc.

Luke Zhang - 6 years, 4 months ago

Log in to reply

@Luke Zhang I agree with u Luke, the solution from Maurice Patel is confusing.. and why the another A.P. series of 200 terms(i.e. 4,9,14,.....999) is not included in answers? In my opinion for n=1/5 and it's multiples,2n+3n is always an integer.So the answer should be 1000.

Bhupendra Jangir - 6 years, 3 months ago

Can u post how to get the inequality?

Luke Zhang - 6 years, 4 months ago

This is the easiest method , isn't it ?

A Former Brilliant Member - 6 years, 4 months ago

Log in to reply

A short cut is that you can show that 5 k 5k , 5 k 2 5k-2 , 5 k 3 5k-3 , 5 k 4 5k-4 is possible by quoting a random value n + { n } n+\left\{ n \right\} , then showing that 5 k 1 5k-1 is skipped entirely due to the fact that for 5 k 5k , n n is an integer.

Julian Poon - 6 years, 4 months ago
Chase McCloskey
Feb 2, 2015

First, I defined the function f(x)=⌊2x⌋+⌊3x⌋ for all real numbers x between 1 and 1000 inclusive. The problem can then be restated as: "Find the range R of f(x), and count the elements in R."

Then I assumed that the range of f(x) consisted of all the integers in the given interval, and tried to find a real number x to satisfy each value for f(x). Experimenting with the first few integers allowed me to deduce that f(x) = 1 when x = 1/3 (say), f(x) = 2 when x = 1/2, and f(x) = 3 when x = 2/3. All efforts to find a value for x so that f(x) = 4, however, were fruitless. I quickly convinced myself that f(x) could never equal 4 by the following (non-rigorous) lines of thought:

If x=1, f(x) = ⌊2(1)⌋+⌊3(1)⌋ = 2+3 = 5.

If x<1 by a very small amount, then ⌊2x⌋=1 by definition of the floor function, and⌊3x⌋ = 2 similarly. Hence, f(x) would equal 1+2, or 3.

I then observed that the same behavior would continue for all integral values of f(x) of the form 5n+4, where n = 0,1,2,... since f(x) is montonically increasing and the range (again, just considering integer values for f(x)) would consist of multiples of 5 if the floor brackets were absent from the 2x and 3x terms. There were 200 such integers in the interval [1,1000]; hence, there must be 1000 - 200 = 800 calculable values for the range of f(x).

Thus, the answer to be sought in the original problem is 800.

Q.E.D.

I really like your answer. I think it's much more clear to not-that-prepared minds like mine :)

Roberto Villadangos Carrera - 6 years, 4 months ago

We could show, for integers k k , that lim ϵ 0 f ( k ϵ ) = 5 k 2 \lim_{\epsilon \to 0} f(k - \epsilon) = 5k - 2 . Combining that with f ( k ) = 5 k f(k) = 5k and f f being increasing shows that there is no value of x x such that f ( x ) = 5 k 1 f(x) = 5k-1 .

Matt McNabb - 6 years, 1 month ago

You will always be my inspiration sir. You have solved almost all hard problems on brilliant which I've seen. Can you please tell me the sources which you refer while learning.

Thanks a lot!

Aryan Gaikwad - 5 years, 11 months ago

Log in to reply

Hello, Aryan!

First of all, thank you for your kind words. I'm flattered and humbled to hear in particular that I and my activities on Brilliant.org have inspired you. I have compiled a vast library of mathematical literature on my computer via free pdf files, so it would be impossible to share everything with you in a text box. I can, however, strongly recommend the following links for exploring mathematical topics, as these are among the most visited bookmarks on my laptop:

http://mathworld.wolfram.com/ http://www.math.niu.edu/Papers/Rusin/known-math/welcome.html https://en.wikipedia.org/wiki/Portal:Mathematics https://proofwiki.org/wiki/Main_Page http://www.mathpages.com/ http://www.recmath.org/ http://www.mathpuzzle.com/ http://math.stackexchange.com/ http://www.archimedes-lab.org/ https://plus.maths.org/content/ http://eqworld.ipmnet.ru/ http://www.numberphile.com/ http://fractalfoundation.org/ http://www.cut-the-knot.org/ http://www.euclideanspace.com/ http://polymathprojects.org/ http://mathigon.org/ https://projecteuler.net/ http://unsolvedproblems.org/ http://www.quadibloc.com/ http://divisbyzero.com/ https://www.khanacademy.org/ https://cp4space.wordpress.com/ http://www.math.utah.edu/~pa/math/conjectures.html

I would also strongly recommend checking out works on recreational mathematics to acquire a sense of the beauty, universality, and fun involved in doing and learning about math. I was introduced to the joys of mathematics and creative problem-solving at 8 years old by reading old essays by Martin Gardner on the subject. I also really enjoy reading the wikis and tutorials written by users of this site; many of the other Brilliant.org members are specialists in a given area and explain the intricacies of that area far better than most, given what I can tell from reading them. You can learn quite a lot just by exploring the site.

Thanks again for your post! Best,

~Chase

Chase McCloskey - 5 years, 11 months ago

Log in to reply

Thanks a lot for your reply. It's greatly appreciated by me. I will surely read through all the links you have provided and I hope one day I'll be able to become a good problem solver, just like you.

~ Aryan

Aryan Gaikwad - 5 years, 11 months ago
Rohit Sachdeva
Apr 14, 2015

Let n=i+f, where i is the integer part & f is the fractional part

[2n]=[2i+2f]=2i+[2f]

[3n]=[3i+3f]=3i+[3f]

[2n]+[3n]= 5i+[2f]+[3f]

Case I: 0 f < 0.333 0≤f<0.333

[2f]=0 & [3f]=0

Case 2: 0.333 f < 0.5 0.333≤f<0.5

[2f]=0 & [3f]=1

Case 3: 0.5 f < 0.667 0.5≤f<0.667

[2f]=1 & [3f]=1

Case 4: 0.667 f < 1 0.667≤f<1

[2f]=1 & [3f]=2

So [2f]+[3f]=0,1,2,3

Hence we have to count numbers of form 5i, 5i+1, 5i+2, 5i+3

Or we can subtract 4, 9, 14, 19,....

Which will count to 200

Hence answer is 800

Stewart Gordon
Jul 27, 2015

Define f ( m ) = m 3 + m 2 f(m) = \left\lfloor \frac{m}{3} \right\rfloor + \left\lfloor \frac{m}{2} \right\rfloor . Obviously this has the same range as the given expression as a function of n n .

Clearly, f ( m ) = f ( m ) f(m) = f(\lfloor m \rfloor) , so we need only consider integer values of m m .

Observe that f ( 0 ) = 0 f ( 1 ) = 0 f ( 2 ) = 1 f ( 3 ) = 2 f ( 4 ) = 3 f ( 5 ) = 3 f ( m + 6 ) = f ( m ) + 5 f(0) = 0 \\ f(1) = 0 \\ f(2) = 1 \\ f(3) = 2 \\ f(4) = 3 \\ f(5) = 3 \\ f(m + 6) = f(m) +\ 5 so that for integer k k , m f ( m ) = k \exists m\ f(m) = k if and only if k ≢ 4 ( m o d 5 ) k \not\equiv 4\ (\mathrm{mod}\ 5) .

Obviously in [ 1 , 1000 ] [1, 1000] , there are 800 \boxed{800} such integers.

Hi friends, after solving this question just for the fun of it I tried a different approach .

I drew the graph of y =floor(2x) + floor(3x) .

I found that for the first 63 natural numbers, there were 50 solutions , therefore for 1000, how much ??? (using Proportionality relations) 793.65 ; from the first 25 solutions I got 781.25 solutions for 1000 natural numbers.

Now for 101 natural numbers I found 80 solutions implying 792.08 solutions for 1000 natural numbers .

Similarly , for 126 natural numbers I got 100 solutions implying 793.65 solutions for 1000 natural numbers .

See, the value progresses steadily towards 800 , isn't it interesting ?

Using graphing it's pretty easy to see that it just skips every 5th number, and looking at the equation it just kinda looks like one that would skip something every n digits. You don't really need to count to more than 5 on the graph to check in my opinion.

Alex Li - 5 years, 1 month ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...