- Search Metals Mine
-
Codebreaker replied Feb 10, 2009Guys Sorry that this thread has been temporarily hijacked by this weirdo!! I have emailed the Admin to get him kicked off the forum. Also, I have sent a link for this thread to the group at Bradford (of course I know them - I did my Post-Doc work ...
Optimized Trend Trading
-
Codebreaker replied Feb 10, 2009Even better - FXTimer has beaten me to it!! Thanks for posting the Russian indy, FXTimer. Guys - take a look at the code. The matrix construction is elegant followed by a simple but effective Gaussian pivoting routine. So the cyclic extension ...
Optimized Trend Trading
-
Codebreaker replied Feb 10, 2009Chamois You are right. I read Despain's paper during my degree and used some of the ideas in my PhD work. It implements nothing more than a bog-standard Cooley-Tukey FFT with the old CORDIC trick, which avoids the use of floating-point ...
Optimized Trend Trading
-
Codebreaker replied Feb 10, 2009Guys FXtimer's post #287 is worth noting. For clarity, the algorithm described by Despain is NOT known as the "VFFT". He simply used the word "very" in front of the words "fast fourier transform" in the paper's title. "The" VFFT was invented by ...
Optimized Trend Trading
-
Codebreaker replied Feb 10, 2009VFFT — Guys Don't get too hung up on this one. Despain's paper was published in 1979 before the advent of the PC and general desktop computing power. As you know, the FFT can be implemented by decimation-in-time or decimation-in-frequency (or ...
Optimized Trend Trading
-
Codebreaker replied Feb 9, 2009That is just Euler's theorem. Remember that exp(i*theta) = cos(theta) + i*sin(theta) where i is the square root of -1. Goertzel references everything to the REAL part which is the cosine. Sine is imaginary!!! So shifted 90 degrees. CB
Optimized Trend Trading
-
Codebreaker replied Feb 9, 2009Yup - that'll fix it. Actually, I avoid the atan() function altogether as it is horrible and use acos() instead as it is always right :-))CB
Optimized Trend Trading
-
Codebreaker replied Feb 8, 2009Wowzer! — Guys I turn my back for an afternoon singing at the Cathedral and you all post at once!! Great! :-)) But it will take me a while to get thru all the questions! All I will say here -- as it is almost 1 in the morning and I have work ...
Optimized Trend Trading
-
Codebreaker replied Feb 8, 2009And here'sthe same statement hopefully containing the equity curve.... CB
Optimized Trend Trading
-
Codebreaker replied Feb 8, 2009Interesting EA ideas? — Guys A friend of mine is quite into commercial EAs. I have always been rather sceptical of them as they always make claims that seem too good to be true. But she recently acquired FAP-Turbo and Piptronic. I have to say ...
Optimized Trend Trading
-
Codebreaker replied Feb 8, 2009Todd Please read the earlier posts in this thread where we discuss the HP filter, theory and limitations. The indy will very likely repaint because the end points change as new data arrives. This can be as many as the last 10 or so bars. So the HP ...
Optimized Trend Trading
-
Codebreaker replied Feb 7, 2009Guys Thanks for the GREAT posts recently - this thread is really hotting up! Anx - the Goertzel indy should really interest readers. I have been using this indy for about a year now and have made many modifications to it to do various things. ...
Optimized Trend Trading
-
Codebreaker replied Feb 7, 2009Thanks for sharing — Hi anx Thanks for sharing these indicators! Just for clarity for readers, the HP_2 indy is NOT a Hodrick-Prescott filter. It is just a very simple weighted linear FIR. But it has some nice properties! Readers - take at ...
Optimized Trend Trading
-
Codebreaker replied Feb 6, 2009Hi The technique I am talking about is described fully in this paper.. url CB
Optimized Trend Trading
-
Codebreaker replied Feb 6, 2009Great work! — Hi Carl Fantastic work!! The chart looks great. And you will have noted that if you remove noise in the optimum way (as you have) you often get a slightly LEADING indication of movement. You are on your way to some great pips!!! ...
Optimized Trend Trading
-
Codebreaker replied Feb 6, 2009Crodzilla - It is quite straightforward. Look at the GX algorithm in post #215. The counter t goes from 0 to last(x) where x is the input price vector. The phase that is output for the desired spectral component is referenced to the START of this ...
Optimized Trend Trading
-
Codebreaker replied Feb 5, 2009Hi Challenger The first interpretation is right. I "bin" the ticks over the period then apply my analysis to the "binned" bars with this rather unconventional TF. This is the real key to FX analysis IMHO. CB
Optimized Trend Trading
-
Codebreaker replied Feb 5, 2009No. The goertzel algorithm will give you absolute phase referenced to the start of the analysis. But be VERY careful WHICH END of your data is the reference. In MT index[0] is the CURRENT bar and index[last] is the OLDEST bar. Some indys have... for ...
Optimized Trend Trading
-
Codebreaker replied Feb 5, 2009Optimum time frames — Hi Challenger You are absolutely right. Some timeframes are definitely better than others for different pairs. If you look at the theory of fractal time series you will see that different "embedding dimensions" (which is ...
Optimized Trend Trading
-
Codebreaker replied Feb 5, 2009No morons allowed on this thread!!! — Ari When I started this thread, we had the usual sprinking of know-nothing morons who just poured scorn. Mercifully, as the thread developed and they saw the quality of the contributions from many good ...
Optimized Trend Trading