What is the best way to deal with the filter coefficients? - page 5
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
It's not exactly advertising. Finware made a program for digital filtering and came up with filter names like FATL, SATL (fast, slow LPF). There were also articles by Kravchuk about the digital filter system. All this aroused a lot of interest and people started to make free alternatives for Finware, so Kenny-Goodman digitalfilters (KG-filters) appeared.Then it was refined and Digital Method Generator appeared, and the names of filters like FATL, KGLP were taken from previous programs.
Here I have shown three low-pass filters - fast, medium, slow - in 3 minutes on my kneehttps://www.mql5.com/ru/forum/172847/page2#comment_4515151
Time to open a company already? ))
Sorry to be amateurish, but are digital filters like SATL FATL etc. close to what you do, or from another field?
I understand it's something in electronics. What is the philosophy and theoretical validity of applying them to quotes?
Indicators like SMA, MACD are digital filters. The SMA is a low pass filter, the MACD is a band pass filter. The SMA is just the average value of N bars. And the average value is used in many areas. The problem is that each area comes up with its own name. The SMA differs from other low-pass filters in parameters like quality of filtering and time lag(more details). As a rule, the better the filtering, the greater the lag (this can be bypassed, but then there would be redrawing). For example, the SMA lags half of its period, for a period of 21 we get a value for 11 bars. That's not bad, it's also fast to count, but in return we sacrifice the quality of the filtering.
Regarding "every area comes up with its own name". Wavelet analysis for example. Lots of water about the wonderful properties of wavelets. But little is it said that a wavelet is essentially a digital filter. That said, you can do wavelet analysis with low-pass filters (the same SMA), but use the filtering parameters you want and not be limited to a small set of wavelets.
Here I have shown three low-pass filters - fast, medium, slow - in 3 minutes on my kneehttps://www.mql5.com/ru/forum/172847/page2#comment_4515151
Time to open a company already? ))
Regarding "every area comes up with its own name."
Well, not every area. Usually it's the accepted terminology, though.
But in the market, in trading, it's very common. You can start anywhere.) It's all abstruse terminology that hides well known methods, most of which are known to a tenth-grader from school.
Well, not all of them. Usually it's the accepted terminology, though.
But in the market, in trading, it's very common. You can start anywhere.) The terminology is very abstruse and hides well-known methods, most of which are known to a tenth-grader from his school course.
Well, they don't study filters at school, and that MA is a bad filter - they hide it altogether
Indicators like SMA, MACD are digital filters. SMA is low pass, MACD is bandpass. In this case SMA is just an average value over N bars. And the average value is used in many areas. The problem is that each area comes up with its own name. The SMA differs from other low-pass filters in parameters like quality of filtering and time lag(more details). As a rule, the better the filtering, the greater the lag (this can be bypassed, but then there would be redrawing). For example, the SMA lags half of its period, for a period of 21 we get a value for 11 bars. That's not bad, it's also fast to count, but in return we sacrifice the quality of the filtering.
Regarding "every area comes up with its own name". Wavelet analysis for example. Lots of water about the wonderful properties of wavelets. But little is it said that a wavelet is essentially a digital filter. That said, you can do wavelet analysis with low-pass filters (the same SMA), but use the filtering parameters you want and not be limited to a small set of wavelets.
Why is the MACD bandpass? The difference of two Exponential MA is taken, it is shown as a histogram (vertical bars), then a slightly modified Simple MA is taken from the difference and shown as a dashed line.
You're wrong about wavelets, I do them. You cannot simulate a wavelet with digital filters. You can very roughly mimic the Fourier transform with bandpass filtering. But FFT is better and faster.
Perhaps I will write my next article about wavelet filtering where delays are minimal and not comparable to FFT.
AF is studied at school, just not called that. MA is neither bad nor good - it is simply one of the simplest filters. Any MA.
Simple MA with length 5 is basically FIR filter with coefficients {0.2, 0.2, 0.2, 0.2, 0.2}
Even a lot, two are enough: trand and oscillator)), and all this is secondary, the main thing is to make more PR and fog)
Oscillator is bandpass or upper-pass filter. Bingo, let's take a bandpass filter, do the math on a crank, and give it to the Market as a mega-graphic super oscillator named after Volchanskiy! )) The main thing is to color the curve sections in canary colours )))