- Search Metals Mine
-
Codebreaker replied Feb 4, 2009Crodzilla - good to have you on the thread! No need to apologise - I can see from your comments and Excel analysis that you already have a good understanding of the topic and are learning fast! It is worth remembering that MetaTrader has a DDE ...
Optimized Trend Trading
-
Codebreaker replied Feb 1, 2009Plots below... — In the plots below in post #205, the orientation is the same. It is just that after processing, I only wanted the 4 dominant components. So I just drag-and-stretched the corner of the 3D plot to a reasonable size to make the ...
Optimized Trend Trading
-
Codebreaker replied Feb 1, 2009Addendum... — One final point that ngawang touched upon... Goertzel HATES a large DC offset in the input data!! It can mess up the spectral estimates. So remove the offset before analysis. You don't need anything fancy. I have tried all sorts ...
Optimized Trend Trading
-
Codebreaker replied Feb 1, 2009Goertzel and Jurik JMA/VEL — Hi Guys You have been busy while my back was turned!! Regards Goertzel, remember that in most books etc they give the most general complex form, ie the version of the algorithm that can handle complex signals ...
Optimized Trend Trading
-
Codebreaker replied Jan 23, 2009Absolutely right! It is vital that there is no phase distortion in the filter as want to add the "signal" bin components back together with the correct amplitude and phase (Goertzels algorithm measures both accurately even in the presence of quite a ...
Optimized Trend Trading
-
Codebreaker replied Jan 22, 2009Ngawang -- the cycles change in period and amplitude quite slowly after requisite signal processing. So it is quite easy to keep track of them. Each "ridge" is a period, slowing changing. You can see that both the period and amplitude (and phase) ...
Optimized Trend Trading
-
Codebreaker replied Jan 22, 2009Thanks for your kind words - we do our best!! To answer your questions... (1) The Goertzel algorithm specifies that you MUST have 3 times the number of points as the maximum period you want to measure. I generally measure cycles up to 120 bars ...
Optimized Trend Trading
-
Codebreaker replied Jan 15, 2009Indeed I do know about this, but it is a big subject! You will need quite a lot of statitical physics to fully understand it. For a start, download this paper and have a read.... linkinghub.elsevier.com/retrieve/pii/S0378437102019088 CB
Optimized Trend Trading
-
Codebreaker replied Jan 15, 2009Of course, I saw this enigmatic answer. I have neither the time nor the inclination to expend any effort on this. One could approach this a hundred different ways. If I get an interested MSc student I might get them to take a look. But I don't think ...
Optimized Trend Trading
-
Codebreaker replied Jan 14, 2009CSSA? — Hi asm Yes I did. I programmed SSA for both Matlab and also in C for my trading platform. Fabulous technique but, like the Hodrick-Prescott filter we have also talked about, the method is non-causal. This means that the right-hand ...
Optimized Trend Trading
-
Codebreaker replied Jan 13, 2009Thank you for the kind comments :-)) Yes and yes. I use the ULLMA to get a smooth price. I then do a spectral analysis based on the Goertzel algorithm and assign dominant cycles to "signal" and smaller amplitude cycles to "noise". Hence we get SNR. ...
Optimized Trend Trading
-
Codebreaker replied Nov 28, 2008It depends what you want to do. Sevchik's method for measuring fractal dimension is technically (mathematically) very accurate but requires a huge amount (theoretically infinite amount) of data so is, in practice, not very useful. If you want to ...
Optimized Trend Trading
-
Codebreaker replied Nov 23, 2008Me too! It is great that some serious folk are getting involved with the forum. I am very much hoping that between us, we can develop some seriously profitable systems. I'll have one of my students look at the Jurik stuff - would make a good Final ...
Optimized Trend Trading
-
Codebreaker replied Nov 23, 2008Please DO NOT pollute this thread with extraneous matters. It is for trend trading and signal processing discussions ONLY. PM folk for private questions! CB
Optimized Trend Trading
-
Codebreaker replied Nov 23, 2008I'm not surprised to hear it. Don't forget an EMA is an IIR with all the nasty phase characteristics that entails! I cannot begin to imagine how hideous the phase response of 19 or so nested EMAs would be... Doesn't even bear thinking about! Phase, ...
Optimized Trend Trading
-
Codebreaker replied Nov 23, 2008JMA is indeed reproducible with simple Nyquist!!! — By a useful coincidence, I have had an email reply from a chap I was corresponding with re Jurik JMA and also Jurik VEL filters. He own the genuine pukka product as a DLL for Excel, C++ etc. ...
Optimized Trend Trading
-
Codebreaker replied Nov 23, 2008Please post the code for jma.mq4 and I'll take a look. I would be VERY surprised if this was the real deal as Jurik is unlikely to have ever released the details of his algorithm - certainly never the source code!!! CB
Optimized Trend Trading
-
Codebreaker replied Nov 20, 2008Hi Kudzu In many ways, a choice as good as any other. But why, precisely, these values? I always like to crunch the maths....... CB
Stochastic secrets
-
Codebreaker replied Nov 20, 2008Hi PipTrees Thanks for the explanation - I can see exactly why you reversed. But I am currently too stupid to have learned this rule. Which (on this lucky occasion) saved me from an unnecessary whipsaw :-)) CB
4H 5EMA System
-
Codebreaker replied Nov 20, 2008Yowza! I thought I had this system just about cracked but then PipTrees throws in this left-ball!! I thought this was simple! In my chart (IG-Index) I tagged 1 the SetUp bar and 2 the Reversal bar. I went short at the Close of the reversal bar at ...
4H 5EMA System