Question about derivative approximation through nested functions

In summary, there are multiple derivative approximation formulas that work for nested functions, with the variable "a" denoting the nesting depth. These formulas approach the derivative of the inner function for a higher value of a, and can also be combined to approach the derivative of a combined inner function. However, in some cases, it may be necessary to subtract 1 from the values and recalculate to get an accurate approximation. This process can also be expressed as a recursive definition of a sequence of functions, with a conjecture that the derivative of the original function can be found by taking the limit of a ratio involving certain constants and the nested functions.
  • #1
Matt Benesi
134
7
Is there a name for these types of derivative approximation?

I've encountered several different derivative approximation formulas, all of which work.

For this series of formulas, the variable "a" denotes the nesting depth (a refers only to the nested part of the equation, not the initial [itex] x - [/itex]):

For this example, a=4: [itex]f(x,n,a)=f(x,n,4)= x - \sqrt[n]{x^{n}-x+\sqrt[n]{x^{n}-x+\sqrt[n]{x^{n}-x+\sqrt[n]{x^{n}-x}}}} [/itex]

This next formulas approach the derivative of the inner functions* when taking [itex]\dfrac{f(...,a)}{f(...,a+1)}[/itex] or sometimes ***[itex]\dfrac{1- \dfrac {f\left(...,a\right)}{f\left(...,a+1\right)}}{1- \dfrac{f\left(...,a+1\right)}{f\left(...,a+2\right)}}[/itex] for higher a. Note that whatever pattern the equation follows, it will continue to follow if you delve deeper into it, but please don't let this aside distract you for now (you can find the derivative in it multiple times by subtracting the derivative from the approximate derivative, dividing this value by a deeper a derivative).

*For the following equation, the inner function is x^n:
[tex]f(x,n,a)=x-\sqrt[n]{x^{n}-x+\sqrt[n]{x^{n}-x+\sqrt[n]{x^{n}-x+\sqrt[n]{x^{n}-x+...}}}}[/tex]
so [itex]\dfrac{f(x,n,a)}{f(x,n,a+1)}\to nx^{n-1}[/itex] OR [itex]\dfrac{1-\dfrac{f(x,n,a)}{f(x,n,a+1)}}{1-\dfrac{f(x,n,a+1)}{f(x,n,a+2)}}\to nx^{n-1}[/itex]

For this one, the inner function is Bx:
[tex]f(x,B,a)=x-\log_B [B^x-x+\log_B [B^x-x+\log_B [B^x-x+\log_B [...]]]][/tex]
so *** [tex]\dfrac{f(x,B,a)}{f(x,B,a+1)}\to B^{x}\ln B[/tex]
as a increases (or for larger B and x).

In fact, all of the basic formulas that use - x + (the repeated formula...) appear to approach the derivative of the inner formula for f(...,a)/f(...,a+1) except in conditions when the functions and inverse functions used have limited well defined domains (such as cosine and cosine-1).

Combining the functions results in approaching the derivative of the combined inner function:
[tex]f(x,n,B,a)=x-\sqrt[n]{\log_B [ B^{x^n}-x+ \sqrt[n] {\log_B [ B^{x^n}-x+ \sqrt[n]{ \log_B [ B^{x^n}-x+... }}} ]]] [/tex]

Note that it is set up to take x^n first, then take B^(x^n) next (as if it were infinitely iterated so that it is algebraically sound). The "[" symbol doesn't show up to clearly under the radical. Anyways...

As with the other -x + ... functions, this one approaches the derivative of the inner function [itex]B^{x^n}[/itex]
***[tex]\dfrac{f(x,n,B,a)}{f(x,n,B,a+1)}\to n\,{x}^{n-1}\,{B}^{{x}^{n}}\,log\left( B\right)[/tex]

***Note that for all of these functions, you can end up with such close values of f(...,a) and f(...,a+1) that you sometimes need to subtract out 1, then calculate the derivative with the new values. You can always tell when this happens, because if you get 1.000...00XXXXX... (XXX's are beginning of different digits) when you divide out f(...,a)/f(...,a+1), it's time to subtract 1 from it, and subtract 1 from the next set f(...,a+1)/f(...,a+2), and then divide the new values out to get your approximate derivative.
 
Physics news on Phys.org
  • #2
Perhaps there will be more comments if we phrase this differently. The word "nested" suggests iteration of a function, but what you are doing is recursively defining a sequence of functions.

The general pattern appears to be:

Let [itex] f(x) [/itex] be a given function with inverse function [itex] f^{-1}(x) [/itex]

Define [itex] f_n(x) [/itex] recursively by:

[itex] f_1(x) = f^{-1}( f(x) - x) [/itex]
[itex] f_n(x) = f^{-1}( f(x) - x - f_{n-1}(x)) [/itex]

A conjecture is that at each point x, the derivative of [itex] f(x) [/itex] is

[itex] f'(x) = \lim_{n \rightarrow \infty} \frac { A + B(x - f_{n+1}(x)) }{A + B(x - f_{n}(x)) } [/itex]

where A and B are some constants that work for all x. In your examples A is 0 or 1 and B is 1 or -1.
 

1. How do you approximate a derivative through nested functions?

In order to approximate a derivative through nested functions, you can use the chain rule. This involves breaking down the nested functions into individual functions and then multiplying their derivatives together.

2. What is the purpose of approximating a derivative through nested functions?

The purpose of approximating a derivative through nested functions is to find the rate of change of a function at a specific point. This can be useful in many scientific fields, such as physics and engineering, where understanding the rate of change is crucial.

3. Can you provide an example of approximating a derivative through nested functions?

One example of approximating a derivative through nested functions is finding the derivative of a composite function, such as sin(x^2). This can be done by using the chain rule and breaking the function down into sin(u) and u^2, where u is equal to x^2.

4. Are there any limitations to approximating a derivative through nested functions?

Yes, there are limitations to approximating a derivative through nested functions. This method may not always give an accurate result, especially if the functions involved are complex or have multiple variables. In these cases, it may be necessary to use other methods, such as numerical differentiation.

5. How does the accuracy of the approximation change with the number of nested functions?

The accuracy of the approximation can vary depending on the number of nested functions. Generally, the more nested functions there are, the less accurate the approximation will be. This is because each additional function adds another layer of complexity and potential error in the calculation. It is important to carefully consider and analyze the nested functions in order to minimize errors in the approximation.

Similar threads

Replies
4
Views
367
Replies
11
Views
2K
Replies
4
Views
885
Replies
3
Views
1K
  • Calculus
Replies
13
Views
1K
Replies
3
Views
1K
Replies
1
Views
939
Replies
5
Views
1K
  • Calculus
Replies
1
Views
1K
Back
Top