|
PureBytes Links
Trading Reference Links
|
At 10:39 PM 8/16/2004, you wrote:
>I hope you will give me a brief explanation of the difference between time
>and frequency filters.
It is sort of complicated mathematically.
Every time-varying signal such as the sound coming out of your stereo system of the price of a stock can be represented as a value vs. time.
> An electronic instrument called an "oscilloscope" can convert
the stereo sound into a picture you can see.
> A real-time data feed and TradeStation charting can convert
the stock price into a picture you can see.
Each such picture is a picture of value (on the vertical axis) vs. time (on the horizontal axis) so this is referred to as the "time domain".
Each signal can also be represented the sum of pure tones - sine waves of different frequencies. The relationship between the two representations is called Fourier Analysis after the guy who figured this out.
<http://www-gap.dcs.st-and.ac.uk/~history/Mathematicians/Fourier.html>
The values of any signal over any window of time can be duplicated exactly by a sum of such sine waves. This is called the "frequency domain" representation.
<http://mathworld.wolfram.com/FourierSeries.html>
As an analogy, consider a photograph of the Mount Everest and a topographic survey map of the same area. Both are different views of the same thing but one or the other will be best for different kinds of analysis.
If you are trading stocks, the time domain is usually more relevant since you care about the money you make vs. time.
If you are designing a stereo system you would usually work in the frequency domain to design the response to low notes (low frequencies) and high notes (high frequencies) with the bass and treble controls. But if you were designing the reverberation effects for the stereo system you would probably work in the time domain since these are time-related echoes.
But you can design filters in either domain. The two domains are simply a different abstract representation of the same thing, just as are the photograph and topographic maps images of Mount Everest.
There are also combinations of the two domain representations. "Wavelets" are sort of a mixture of the time and frequency domains - see:
<http://www.amara.com/IEEEwave/IEEEwavelet.html>
Hope this helps.
Bob Fulks
|