Trading Strategies Based On Digital Filters - page 48
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
Dvarrin,
A few questions.
1. How did you create your RSTL? That doesn't look like any RSTL i have ever created.
2. Trading the turn of STLM isn't the correct way to use the digital filters IMHO. STLM+SATL defines the "direction" that you trade in.
There are just a lot of "rules" with regards to these particular filters that are in the ATCF document in this thread that I think are worth reading. No matter what strategy you trade, I just think it's important to know the intentions of certain indicators (see my comment about CCI).
clI was using the method described in this thread. I first created the SATL and I tried to reduce the number of coefficients without adding to much noise. So I was using 11 coefficient for the SATL. Then for RSTL I used 11 + 11/2 = 16 or 17 coeficients. and I got that result when I pu both on the chart. Did I do something wrong?
I was using the method described in this thread. I first created the SATL and I tried to reduce the number of coefficients without adding to much noise. So I was using 11 coefficient for the SATL. Then for RSTL I used 11 + 11/2 = 16 or 17 coeficients. and I got that result when I pu both on the chart. Did I do something wrong?
That's probably my fault a bit. I seem to remember posting a long time ago that I didn't like SATL's with a large number of coefficients. After working with ATCF months ago, i came to realize that wasn't the most important thing (at least for me).
Delay your SATL by half of P. So, if you have a 70 50 filter, Delay it by 35.
Sorry for any confusion,
cl
I was using the method described in this thread. I first created the SATL and I tried to reduce the number of coefficients without adding to much noise. So I was using 11 coefficient for the SATL. Then for RSTL I used 11 + 11/2 = 16 or 17 coeficients. and I got that result when I pu both on the chart. Did I do something wrong?
The default indicators you can download are "fair". I think they can be tuned to be much better.
cl
That's probably my fault a bit. I seem to remember posting a long time ago that I didn't like SATL's with a large number of coefficients. After working with ATCF months ago, i came to realize that wasn't the most important thing (at least for me).
Delay your SATL by half of P. So, if you have a 70 50 filter, Delay it by 35.
Sorry for any confusion,
clAnd what do we do with all what Simba explained about the number of coefficients? Was that wrong? So we simply delay the RSTL by half the P we use for the SATL and that's it? That means what we don't nead to care on how many coefficient our SATL or RSTL have?
This would sound better to me and more understandable ;-)
What is better to use for D1 if we have peaks at 18, 42 and 79? I use 79 for P1, but if I use 18 for D1, then the STLM will be less smooth than with 42 for D1. What is the trading entry signal? How should we use STLM exactly? we use the crossing of the 0 level? We entry when the cross occurs and according to Hurst we know that price will go twice as far?
Thank you very much again :-)
non stationary CHIRP
Here is non stationary CHIRP signal as .HST file. It changes frequency and amplitude at the same time, you can use it for test of your indicators to simulate non stationary market. See also MESA spectrum of it, amplitude of spectrum is very low and compare to GOLD15 with clear cycle of sin and cos. Description how to use .HST is at post 361 and below
Krzysztof
And what do we do with all what Simba explained about the number of coefficients? Was that wrong? So we simply delay the RSTL by half the P we use for the SATL and that's it? That means what we don't nead to care on how many coefficient our SATL or RSTL have?
This would sound better to me and more understandable ;-)
What is better to use for D1 if we have peaks at 18, 42 and 79? I use 79 for P1, but if I use 18 for D1, then the STLM will be less smooth than with 42 for D1. What is the trading entry signal? How should we use STLM exactly? we use the crossing of the 0 level? We entry when the cross occurs and according to Hurst we know that price will go twice as far?
Thank you very much again :-)No, Simba was not wrong at all. He is right. If you have 60 coefficients your period it technically 60 (those 60 coefficients are multiplied by the last 60 bars....just like a 60 period SMA...but they have all different weights...and all add up to 1). They way i use to run my SATL's (and still do) is delay them by half of P. I'm trying to go of memory, but if i remember right, the RFTL doesn't have the half period relationship that the SATL/STLM did. I always interpreted the help file as 1/2 of P. That's why i messed with both and found that i personally liked 1/2 of P better. (i use the same P and D method for every pair in terms of how i derive each....so i think it's important to note that....i'm just not picking P and D values that just "look good"). This is simply how I have done it.
A lot of my SATL's had 200 coefficients. Could i get a similar line with less coefficients? I probalby could. I just used the same method over and over to calculate them though. So, it's "mechanical" in how i created the filters and i'm comfortable with the output i got back.
As for the signals, I would read the ATCF article. It discusses the signals in depth. I'm sure everyone trades this differently. It was a long time ago for me :-). These filters are not my main setup these days.
I think it's important to define your method and make it repeatable. And FYI, Simba is a great teacher - take what he says to heart!
Best,
cl
No, Simba was not wrong at all. He is right. If you have 60 coefficients your period it technically 60 (those 60 coefficients are multiplied by the last 60 bars....just like a 60 period SMA...but they have all different weights...and all add up to 1). They way i use to run my SATL's (and still do) is delay them by half of P. I'm trying to go of memory, but if i remember right, the RFTL doesn't have the half period relationship that the SATL/STLM did. I always interpreted the help file as 1/2 of P. That's why i messed with both and found that i personally liked 1/2 of P better. (i use the same P and D method for every pair in terms of how i derive each....so i think it's important to note that....i'm just not picking P and D values that just "look good"). This is simply how I have done it.
A lot of my SATL's had 200 coefficients. Could i get a similar line with less coefficients? I probalby could. I just used the same method over and over to calculate them though. So, it's "mechanical" in how i created the filters and i'm comfortable with the output i got back.
As for the signals, I would read the ATCF article. It discusses the signals in depth. I'm sure everyone trades this differently. It was a long time ago for me :-). These filters are not my main setup these days.
I think it's important to define your method and make it repeatable. And FYI, Simba is a great teacher - take what he says to heart!
Best,
clI don't understand. You delay the SATL? I thought it was the RSTL which is delayed. Could you share which setup you're using today?
Only to resume. If I want to create the filters SATL, RSTL and STLM with peaks at 20, 40, and 80.
For SATL I will use P1=80 and D1=20 for example.
For RSTL I will use P1=80 and D1=20 and delay it by P+/2 = 40
And for STLM I'll use the same coeefficients as for SATL and RSTL to replace in the normal STLM filter.
Is it right?
How do we compute the Nyquist interval? I've seen we have to use it for the delay of the RSTL digital filter.
I don't understand. You delay the SATL? I thought it was the RSTL which is delayed. Could you share which setup you're using today?
Only to resume. If I want to create the filters SATL, RSTL and STLM with peaks at 20, 40, and 80.
For SATL I will use P1=80 and D1=20 for example.
For RSTL I will use P1=80 and D1=20 and delay it by P+/2 = 40
And for STLM I'll use the same coeefficients as for SATL and RSTL to replace in the normal STLM filter.
Is it right?
How do we compute the Nyquist interval? I've seen we have to use it for the delay of the RSTL digital filter.RSTL = SATL delayed by 1/2 of the period or P (like i explained above.....it appears there are a couple of ways to do it....both i have personally used and seen work).
Let's say you have peaks at 20,40,80. For SATL, i want to pass the longer cycles. So, i want to set my P less than 40 (so i capture 40 and 80). I think there are different ways to decide what kind of D to use. Some use a trough (i have done this...works fine) and some use a factor of P(i have done this as well, and it works fine).
STLM = SATL-RSTL. Simply replace them yes.
The ATCF document will go into detail about these filters. It's translated from Russian, but still worth an intense read.
cl
Thank you very much clahn04.
I did everything wrong :-((
To be more precise, when you want to pass the longer cycles for SATL, you take a smaller P1 than 40 in this case, but do you take peak or a valley?
Where can I get that ATCF document?The ATCF document is on this thread. Take a valley.
cl
RSTL = SATL delayed by 1/2 of the period or P (like i explained above.....it appears there are a couple of ways to do it....both i have personally used and seen work).
Let's say you have peaks at 20,40,80. For SATL, i want to pass the longer cycles. So, i want to set my P less than 40 (so i capture 40 and 80). I think there are different ways to decide what kind of D to use. Some use a trough (i have done this...works fine) and some use a factor of P(i have done this as well, and it works fine).
STLM = SATL-RSTL. Simply replace them yes.
The ATCF document will go into detail about these filters. It's translated from Russian, but still worth an intense read.
clThank you very much clahn04.
I did everything wrong :-((
To be more precise, when you want to pass the longer cycles for SATL, you take a smaller P1 than 40 in this case, but do you take peak or a valley? How do you determine which factor of P1 is best for D1?