a trading strategy based on Elliott Wave Theory - page 111

New comment
 
And this is what I was talking about.

Here is Rosh's picture, the quotes are almost the same

And here's mine, at first everything seems fine too, then quotes start lagging and then overtake, maybe a glitch in graphics editor?
 
By the way, you can clearly see how the levels differ between the Student and normal distribution for relatively short channels.
 
ЗЫ А вы на данных какого сервера рисунки приводили последние?


Your quotes are normally superimposed on Vladislav's quotes, but mine do not want to (the same pattern but the time scale is different in different parts), it seems my broker is cheating :), or maybe it's because of the strange period of 31, but I do not know how to quickly get data up to a certain date, I receive using the script StepByStep



I posted using MQ data (I understand it suits you better). I "killed" the data before a certain date simply - manually through the quotes archive. Select the bar and "Delete". I forgot about the StepByStep script, I just haven't tried it myself.

 
By the way, you can clearly see how the levels differ between the Student and normal distributions for relatively short channels. <br / translate="no">


Tell me more - how is it visible? Who has a conditional, and who has normal. I did not immediately see (and not immediately after
I also could not understand it.)
 
(I understand they are more suitable for you).


It's just that they are available to everyone.
 
:) how did I violate them? Reference to the source is, but how to count the variance was my problem, I'm guilty only of not thinking that N for you and for me a little different things, I understand that generally accepted N is the number of degrees of freedom (number of measurements, trials, outcomes, etc.), and for me N is the number of bars counting from 0. Bar 0, and measurements 1, here we have a disagreement :).


OK. After all +/- 1 is nothing compared to the revolution (in autotrading Forex). :-)
By the way, I'm not the author of this formula. It predates me, in the 18th century.
 
How does it show? Who has a condition, and who has a normal one. I didn't see it right away (and I couldn't understand it either after <br/ translate="no"> this phrase).


I use a normal distribution, and Vladislav I understand the Student distribution. Where my picture is superimposed the biggest channel is almost the same not only by the centre line but also by the confidence intervals, while the shorter channels are different, though maybe it seems to me, because the centre lines don't coincide either :(
 
OK. After all +/- 1 is nothing compared to the revolution (in forex auto-trading). :-)


What is interesting, I also thought so before, and you try to replace N with N+1 or N-1, I can't find any channels from this replacement. So, although 1 is a small number but the contribution is significant (to the revolution) :)
 
What is interesting I also thought so before, but you try to change N to N+1 or N-1, from this replacement and I do not find channels. So even though 1 is a small number, it's a significant contribution (to the revolution) :)

Actually, it shouldn't be. A revolution must be stable with respect to slight changes in the number of revolutionaries. Does this happen on all channels or only on short channels?
And in that conclusion I sent you, N is the number of bars on which the calculation is done.
I neglected from the beginning the difference between the number of degrees of freedom and this N.
 
Actually, how it worked, at first I also divided by N as you say and I already thought I was going completely crazy (I pored over the code for 5 hours, even asked you for the output :) ) the line was centered, but the interval boundaries were somewhere in the sky and underground, and only then I thought, that I take the bar number as a variable N in the calculation, but it differs by one from the number of measurements. When I substitute +1 it shows boundaries where it should be (maybe it was Lenin :)))))) ). Maybe it's just a specificity of calculation algorithm.

In any case, even if there is now an error, the figures I posted show that it can be avoided.

N is the number of bars on which the calculation is done. <br / translate="no"> I neglected from the beginning to distinguish the number of degrees of freedom from this N.


I mean the number of degrees of freedom and the number of bars and number of measurements is the same, but the number of bars counts from 0, that's the difference.
1...104105106107108109110111112113114115116117118...309
New comment