Market etiquette or good manners in a minefield - page 54
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
I don't use Matcad, but have seen it here:
http://webmegapolis.com/soft/
http://magnit.ca/2009/01/14/
>> thank you! I didn't find him there, but at least I got to see the girls... -:)
>> get 2001i Pro, you'll be fine.
>> okay! I'm off to find it.
Sergey, once again I tell you - (H+L)/2 is an averaging of the original quotient. The resulting BP (which is from (H+L)/2) has a positive correlation between neighboring samples in the first difference series, which is why it is easily predicted by any method, including your long-born, ingenious and God knows what method. The same series has, as a consequence, inevitable phase delay, which nullifies all the advantages of its easy predictability.
You're engaging in pseudoscientific masturbation and I'm getting tired of telling you that! It's like predicting a muwink - it predicts well, but lags just as well. You can, of course, increase the forecast horizon, but the accuracy drops exponentially with all the wet stuff that comes with it.
You're running in circles and already bores me with this.
What is the phase delay of (H+L)/2? In relation to what? What is the high correlation between ((H[n]+L[n])/2 - (H[n-1]+L[n-1])/2). How is my future bar different from yours? I wonder, how is mine worse?
Here is the series (H[n]+L[n])/2 - (H[n-1]+L[n-1])/2, which I predict (example at a glance):
Here is your favourite ACF
What is the high correlation here? Well, yes, sometimes it may reach 0.3-0.4 on the first bar. Rarely, but it is not a high value at all. And you do not even understand that if there is no good correlation, no network, much less a single layer will not predict, You seem to not understand it at all.
Are you out of your fucking mind, you're calling me names again (you're the first to do so). Ah...I see, so you got tired of ananism and decided to get yourself a lady in the form of NS.
PS: "You're running around in circles and you're pissing me off."
Okay, okay. I won't elaborate on how you're running around, but if you are, I'll leave, or else I'll tell you to "fuck off" and that's not cultured at all. Good luck, at least with your girlfriend.
paralocus 09.06.2009 13:01
to grasn... the temperature of the discussion seems to be rising. It's times like these when you realize how good it is to be stupid and small (I mean me, if you haven't understood). When I will have time for something else, other than studies and trading - I'll start a thread on the psychology of communication between traders and the results of not paying attention to this in trade. Have no doubt, peasants, many of you it will not be useless, especially since the afftar (future) - a well-known person (in small circles, of course ... -:) i.e. duXtor no longer help. Medicine is powerless as they say)
Keep an eye on him, because I think he gets the lady and instead of fucking her, he keeps looking at her and anan.... Jesus, what the fuck he does, I mean NS, he don't talk much about the rest, just scraps of it.
Thank you! I didn't find him there, but at least I got to see the girls... -:)
Sorry, I didn't check the links and gave the wrong ones from memory, I got old: http://lanzone.org.ua/index.php?do=cat&category=soft_cad
So... while you're thinking about the question (and given the time difference with Novosibirsk, you're probably asleep), I decided for the sake of curiosity to try to predict the colour of the bar using the AR method. I tried to predict "lob" and then I tried to predict the bar colour. I tried to predict "directly" direction ("+" or "-") of x[n]-x[n-1] (H+L)/2 without model identification. Similarly, as I expected, it is rubbish, because one cannot do it at once. But I remembered one old idea of series processing and obtained any experimental result (on 15 min of EURUSD 5 000 samples):
The most disappointing thing is that there is no error... But you're right, Serega. Knowing the "direction" in a bar and the mean square "bar breathing" (for some mysticism) we can build a good strategy. So what are your results? You were estimating how much you were letting the system lie, weren't you?
I can predict like that too, by the way. Even without an AR. :) But it didn't give me anything. I can guess "where" the price will go with an accuracy of 80% but I have no PROFITS. It's sad. ;)
Gentlemen, my Matcad is glitching.
Can you be more specific about the build number? And what does the glitch look like? Maybe there's a test case that reproduces the glitch? Apologies for the offtop, just would like to rule out something like this for myself.
Here's your favourite ACF.
What high correlation is there? Well, yes, sometimes it can reach on the first leg like 0.3-0.4. Rarely, but it's not a high value at all. And you don't even understand that if there is no good correlation, no network, much less a single layer will NOT PREVENT, You don't seem to understand it at all.
What a phrase!
Where did you see me talking about ACF? There's no need to make things up and answer yourself with your own drawings.
For you (a tank man) I'll say it again. Read my lips:
You have a strong positive correlation between the adjacent readings of the series (H+L)/2 and it has nothing to do with the quote that you need to predict (in the end) to make a profit. Sergey, where, exactly, do you not understand? You don't know what a series of first differences is or you don't understand what adjacent reads are or maybe you cannot attribute it all to different timeframes?
Let me draw you this for different TFs of EURUSD:
Did you call this a weak correlation? To predict such a series it is enough to know the sign and amplitude of the last reading and that's it! You're not going to profit from it, but you've been making such a big fuss for several years already - Euler, dammit.
I won't even talk about the inevitable federal law - I'm sick of it, as performed for you. You can ask Privalych, he will explain it to you in simple language, like for a tank driver :-). He's patient enough!
The files below so you wouldn't ask me "smart" questions.
I, by the way, am just as good at predicting. And even without any AR. :) But it didn't do me any good. I can guess "where" the price will go with an accuracy of 80% - but I have no PROFITS. It's sad. ;)
You'd better explain to this nugget how it happens, because I am powerless.
Sorry, didn't check the links and gave the wrong ones from memory, got old: http://lanzone.org.ua/index.php?do=cat&category=soft_cad
Thanks anyway -:)
To eire I split the code by dumping the extra functions (cotier reading, tangents, etc.) into a separate file. And the grid stopped learning normally. I've had troubles with it before, but I blamed it on my own mistakes, but this time I decided to check it. It turned out that the code works correctly only if it is gathered in one listing. Here it is before division:
And after:
to Neutron found 2001i Pro. >> I'm putting it on.
You'd better explain to this nugget why this happens, because I'm powerless.
No, no, no, no. I'm not gonna get involved in your arguments. You guys have been fighting for a long time, so I'll stay out of it. Besides, the thing is, you're both a little less than right.