- #36
Ivan Seeking
Staff Emeritus
Science Advisor
Gold Member
- 8,178
- 2,054
zoobyshoe said:No, since I don't know what a "curve fitting technique" is.
We can use some tricks in math - for one, the Fourier expansion - to reproduce nearly any empirically derived curve.
Fourier theorem: A mathematical theorem stating that any function may be resolved into sine and cosine terms with known amplitudes and phases - [provided function meets certain [the Dirichlet] conditions.
This typically produces an infinite series of sin and cos terms, each with less and less significance [decreasing amplitudes], that approximates the empirical curve as precisely as we like - depending on how many terms that we use. I for one use this common technique to produce useful approximations of empirically derived [data] curves, or to approximate other mathematical expressions that can't solved directly by the algorithms available [industrial computers].
I'm not sure why, but assuming you mean it is too old for them to have been able to study it accurately by todays standards,
I meant that this definitely predates any complete solutions...according to what I read.
In fact, the more I think about this the surer I am. I am going to check my bookshelves and see if I can find my source. I am thinking that this may have even been in Feynman's lectures.