The Student Room Group

Is there a rational number between any two irrationals?

Question in the title.

I feel intuitively the answer should be yes, but can't really seem to prove this... Any help?
Reply 1
Yes, all you have to do is look at the digits of both irrational numbers until there is a difference between two digits, truncate the irrationals after this digit (so you have a rational number) add the two rational numbers together, divide by two and you've got a rational number between the two irrationals.

So for example, suppose you've got the two following irrational numbers α,β\alpha,\beta and wlog write them as

α=0.a1a2a3...\alpha = 0.a_{1}a_{2}a_{3}...

β=0.b1b2b3...\beta = 0.b_{1}b_{2}b_{3}...

Where the ai,bia_i, b_i are the digits.

The reason we can say they're between 0 and 1 is because it doesn't matter what the digits are to the left of the decimal point, it's just extra notation if we add them in.

Let's say that ai=bia_i = b_i for 1in11 \leq i \leq n-1, so what we're saying is there is a difference between the two irrationals at the digit in place ana_n

So denote the rational numbers

α=0.a1a2a3...an\alpha' = 0.a_{1}a_{2}a_{3}...a_{n}

β=0.b1b2b3...bn\beta' = 0.b_{1}b_{2}b_{3}...b_n

Then the number α+β2\dfrac{\alpha' + \beta'}{2} is also rational, but note that

α+β2=0.a1a2a3...an+0.b1b2b3...bn2\dfrac{\alpha' + \beta'}{2} = \dfrac{0.a_{1}a_{2}a_{3}...a_{n} + 0.b_{1}b_{2}b_{3}...b_n}{2}

This number will have the first n2n-2 digits identical to the digits of α,β\alpha',\beta' but the last two digits will be such that a<α+β2<ba < \dfrac{\alpha' + \beta'}{2} < b
(edited 11 years ago)
Reply 2
Yes, there is.

Hint for the proof: consider the decimal expansions of the numbers.
Reply 3
Original post by Maths Failure
Question in the title.

I feel intuitively the answer should be yes, but can't really seem to prove this... Any help?


Yes

Edit : as Noble said :smile:

Edit : even more what Noble said :smile:
(edited 11 years ago)
Reply 4
Oh ok thanks guys
Reply 5
Original post by Maths Failure
Oh ok thanks guys


I've just updated my reply with a more mathematical explanation which you might find more helpful in trying to understand.
Reply 6
Original post by TenOfThem
Yes

Edit : as Noble said :smile:

Edit : even more what Noble said :smile:


:eek: I think you accidentally negged me, with that god awful -8 reputation power.

Either that or you're getting your own back for me finding your facebook picture the other day :lol:
Reply 7
Original post by Noble.
Yes, all you have to do is look at the digits of both irrational numbers until there is a difference between two digits, truncate the irrationals after this digit (so you have a rational number) add the two rational numbers together, divide by two and you've got a rational number between the two irrationals.

So for example, suppose you've got the two following irrational numbers α,β\alpha,\beta and wlog write them as

α=0.a1a2a3...\alpha = 0.a_{1}a_{2}a_{3}...

β=0.b1b2b3...\beta = 0.b_{1}b_{2}b_{3}...

Where the ai,bia_i, b_i are the digits.

The reason we can say they're between 0 and 1 is because it doesn't matter what the digits are to the left of the decimal point, it's just extra notation if we add them in.

Let's say that ai=bia_i = b_i for 1in11 \leq i \leq n-1, so what we're saying is there is a difference between the two irrationals at the digit in place ana_n

So denote the rational numbers

α=0.a1a2a3...an\alpha' = 0.a_{1}a_{2}a_{3}...a_{n}

β=0.b1b2b3...bn\beta' = 0.b_{1}b_{2}b_{3}...b_n

Then the number α+β2\dfrac{\alpha' + \beta'}{2} is also rational, but note that

α+β2=0.a1a2a3...an+0.b1b2b3...bn2\dfrac{\alpha' + \beta'}{2} = \dfrac{0.a_{1}a_{2}a_{3}...a_{n} + 0.b_{1}b_{2}b_{3}...b_n}{2}

This number will have the first n2n-2 digits identical to the digits of α,β\alpha',\beta' but the last two digits will be such that a<α+β2<ba < \dfrac{\alpha' + \beta'}{2} < b


Why does ii have to be bounded? If ii was infinite could an argument be made that α\alpha and β\beta are still possibly different numbers?
Reply 8
Original post by Maths Failure
Why does ii have to be bounded? If ii was infinite could an argument be made that α\alpha and β\beta are still possibly different numbers?


No, if the numbers do not differ for some finite ii then they are the same number. You can argue it in a similar way to how 0.999...=10.999... = 1 using infinite series and sequences because if the digits do not differ for any finite ii then the infinite sum of both are exactly the same.
(edited 11 years ago)
Reply 9
Original post by Noble.
:eek: I think you accidentally negged me, with that god awful -8 reputation power.

Either that or you're getting your own back for me finding your facebook picture the other day :lol:


Balls

Let me see if I can redress

xx

I will come back and personal rep tomorrow but I am out today

Sowwy
Reply 10
Original post by TenOfThem
Balls

Let me see if I can redress

xx


Haha don't worry about it, I've got plenty of rep. :colone:
Original post by Noble.
Haha don't worry about it, I've got plenty of rep. :colone:


You call a +6 "plenty"

Hah



















[obviously I have more cos i am soooooooo much older]
Reply 12
Original post by TenOfThem
You call a +6 "plenty"

Hah



















[obviously I have more cos i am soooooooo much older]


Haha, I've just seen you joined this site in Sep 2011 and have 15.5k posts - I think we can safely conclude you like helping people too much, what with you being a maths teacher. You and Mr M should start some Avengers like team of maths teachers.
Original post by Noble.
Haha don't worry about it, I've got plenty of rep. :colone:


Attempted to redress balance :biggrin:

Quick Reply

Latest