How Does Calculus Help Solve Patterns?

  • Thread starter Thread starter lenfromkits
  • Start date Start date
  • Tags Tags
    Calculus
AI Thread Summary
The discussion centers on demonstrating the approximation \(\sqrt{n+1} \approx \frac{n}{2} + 1\) when \(n\) is small. A Taylor expansion around zero is suggested, leading to a linear approximation of the function \(\sqrt{n+1}\). The formula for linear approximation is clarified as \(L(n)=f'(n_0)(n-n_0)+f(n_0)\), which helps in deriving the approximation. The conversation also touches on using the Binomial Theorem for further simplification. Overall, the participants aim to deepen their understanding of these mathematical concepts and their applications.
lenfromkits
Messages
106
Reaction score
0
I need to show that this pattern is true.

\sqrt{n+1} \approx \frac{n}{2} + 1 \ \ when \ n << 1

the smaller 'n' is, the more this is true.

(i.e. for example, \sqrt{1.0000000000001234} \approx \frac{.0000000000001234}{2} + 1 )

Can anyone help?

(this is not a homework assignment. I need it for work)

Thanks.
 
Last edited:
Physics news on Phys.org
You can do a Taylor expand around 0 of \sqrt{x+1}

It is:
\sqrt{x+1}=1+x\left[\frac{1}{2\sqrt{x+1}}\right]_{x=0}+...
 
The linear approximation of f(n) about n_0 is given by:

L(n)=f'(n_0)(n-n_0)+f(n_0) \approx f(n)If f(n)=\sqrt{n+1}, then the linear approximation of f(n) around 0 is:

L(n)~=~f'(0)(n-0)+f(0)~ =~ \frac{1}{2\sqrt{0+1}}(n-0)+\sqrt{0+1}~ =~ \frac{n}{2}+1
 
Thanks! I can see that you have solved my problem. Now I just need to understand 90% more than I understand now :)

Can I please ask where this comes from:

L(n)=f'(n_0)(n-n_0)+f(n_0) \approx f(n)

Thanks,
Len


elfmotat said:
The linear approximation of f(n) about n_0 is given by:

L(n)=f'(n_0)(n-n_0)+f(n_0) \approx f(n)


If f(n)=\sqrt{n+1}, then the linear approximation of f(n) around 0 is:

L(n)~=~f'(0)(n-0)+f(0)~ =~ \frac{1}{2\sqrt{0+1}}(n-0)+\sqrt{0+1}~ =~ \frac{n}{2}+1
 
It's called linear approximation. Approximate your function with a line whose slope is the same as the original function's slope at x=0.
You find this slope from the derivative.
 
Even simpler, since

n + 1 \approx n + 1 + \frac {n^2} {4} = (n+\frac{1}{2})^2

if n<<1


\sqrt {n+1} \approx n + \frac{1}{2}
 
lenfromkits said:
Thanks! I can see that you have solved my problem. Now I just need to understand 90% more than I understand now :)

Can I please ask where this comes from:

L(n)=f&#039;(n_0)(n-n_0)+f(n_0) \approx f(n)

Thanks,
Len

You use a tangent line to approximate the function at n_0. The equation of the tangent line is obviously given by y=f&#039;(x_0)(x-x_0)+f(x_0).

http://www.function-grapher.com/graphs/fc544b3a572de466f056ae4620193c3f.png
 
Last edited by a moderator:
Back
Top