Find the Approximate Value of this Cube Root

  • Thread starter Thread starter littlemathquark
  • Start date Start date
  • Tags Tags
    Roots
AI Thread Summary
The discussion revolves around approximating the cube root of 63 using various mathematical methods, including Bernoulli's inequality and Taylor expansion. Participants analyze a specific approximation technique that involves rewriting the cube root as a function of 64, leading to an expression that simplifies the calculation. There is confusion regarding the relevance of certain terms in the denominator and the accuracy of the approximation, with error estimates provided. The conversation also touches on the applicability of Halley's method, which is ultimately dismissed. The discussion highlights the nuances of mathematical approximation techniques and their implications for accuracy.
littlemathquark
Messages
204
Reaction score
26
Homework Statement
Find approximate value of ##\sqrt[3] 63##
Relevant Equations
Find approximate value of ##\sqrt[3] 63##
I know a lot of solution of that problem; for example Newton-Ralphson or Taylor expansion.But I saw a solution but I don't understand its theory. The solution like this: ##\sqrt[3]{63}=\sqrt[3]{64-1}\approx 4-\dfrac{4}{3.4^3-1}\approx 3.97905759162...##
 
Last edited:
Physics news on Phys.org
What is ##3.4^3##? Bernoulli is ##(1+x)^r \approx 1+rx## which yields in this case
$$
\sqrt[3]{1-\dfrac{1}{4^3}}\approx 1- \dfrac{1}{3\cdot 4^3}
$$
 
Last edited:
fresh_42 said:
What is ##3.4^3##? Bernoulli is ##(1+x)^r \approx 1+rx## which yields in this case
$$
\sqrt[3]{1-\dfrac{1}{4^3}}\approx 1- \dfrac{1}{3\cdot 4^3}
$$
I don't know. According to your solution ##63^{1/3}\approx 4-1/48\approx 3.979166##
 
littlemathquark said:
I don't know. According to your solution ##63^{1/3}\approx 4-1/48\approx 3.979166##
Forget the factor ##4## on both sides. Bernoulli's inequality says for numbers greater than ##-1##
\begin{align*}
0,99476430197409796491613199693719&\approx \dfrac{1}{4}\sqrt[3]{63}=\left(1-\dfrac{1}{64}\right)^{\frac{1}{3}}\leq 1- \dfrac{1}{3}\cdot \dfrac{1}{64}\\&\approx 0,99479166666666666666666666666667
\end{align*}
with an error of about ##2,73647 \cdot 10^{-5}.## If we use ##\dfrac{1}{3\cdot 4^3 -1}=\dfrac{1}{191}## instead, then we get ##0,99476439790575916230366492146597## which varies by ##10^{-7}.## It is slightly better but not significantly. So I don't know where the irrelevant ##-1## in the denominator comes from, but the rest is Bernoulli's inequality.
 
fresh_42 said:
Forget the factor ##4## on both sides. Bernoulli's inequality says for numbers greater than ##-1##
\begin{align*}
0,99476430197409796491613199693719&\approx \dfrac{1}{4}\sqrt[3]{63}=\left(1-\dfrac{1}{64}\right)^{\frac{1}{3}}\leq 1- \dfrac{1}{3}\cdot \dfrac{1}{64}\\&\approx 0,99479166666666666666666666666667
\end{align*}
with an error of about ##2,73647 \cdot 10^{-5}.## If we use ##\dfrac{1}{3\cdot 4^3 -1}=\dfrac{1}{191}## instead, then we get ##0,99476439790575916230366492146597## which varies by ##10^{-7}.## It is slightly better but not significantly. So I don't know where the irrelevant ##-1## in the denominator comes from, but the rest is Bernoulli's inequality.
Can be used Halley Method?
 
littlemathquark said:
Can be used Halley Method?
No.

It is Maclaurin of ##(1-x)^r## up to the second derivative with an error of ##\dfrac{1}{9\cdot 4^6(3\cdot 4^3-1)}+O(4^{-27})=O(4^{-11})=O(10^{-7}).##
 
Last edited:
littlemathquark said:
I don't know. According to your solution ##63^{1/3}\approx 4-1/48\approx 3.979166##
The step that was omitted is this:

##63^{1/3} = (64 - 1)^{1/3} = 64^{1/3}(1 - \frac 1{4^3})^{1/3} = 4(1 - \frac 1{4^3})^{1/3}##
In the final expression above he's working only with the ##(1 - \frac 1{4^3})^{1/3}## part.
 
Back
Top