Sultonov differential indicator - page 21

 
Yousufkhodja Sultonov:
Posted. Feel free to share your impressions here as well. The code is incomplete, we are looking for the cause of unauthorised redrawing.

Thank you )

 
Dmitriy Skub:

Be thankful for criticism - it helps us develop. What's the use of giving thanks for praise?

One successful prediction is no different from guessing on a coin. Yusuf, you had a cloud of signals on your theories. Where are they all?

Dear Dimitri, I present to your attention the "third" forecast. 4 days ago, on 08.09. at 13-50 https://www.mql5.com/ru/forum/214529/page19I fixed the indicator readings with a certainty that now the price will fall and we are still convinced of the correctness of the indicator forecast - the price has fallen more than 110 points. (4 signs) and continues to fall as the indicator is now showing. In total, over the last week, it has confidently taken about 300 pips on buys and sells on TF M1, a period of 1000 minute bars.
Дифференциальный индикатор Султонова
Дифференциальный индикатор Султонова
  • 2017.09.07
  • www.mql5.com
Уважаемые программисты...
 
Artyom Trishkin:

I'm very busy with my own business at the moment. But I'll find time to have a look during the week.

(chuckles): Thank you.
 
Yousufkhodja Sultonov:
Thank you.

Yusuf, answer me too! No possibility to test, when working in test mode the indicator gives this picture after 100 bars and until the end of the test:


 
Yousufkhodja Sultonov:
Dear Dimitri, I present to your attention the "third" prediction. 4 days ago, on 08.09. at 13-50 https://www.mql5.com/ru/forum/214529/page19I fixed the indicator readings with a certainty that the price will fall and we are still convinced of the correctness of the indicator forecast - the price has fallen more than 110 points. (4 signs) and continues to fall as the indicator is now showing. In total, over the last week, it has confidently taken about 300 pips on buys and sells on TF M1, a period of 1000 minute bars.

You know, back in the day there was an interesting article about scientific research in Tekhnika Molodezhdy magazine. Some scientist announced the discovery of some rays and decided to show the experiment to some other scientist. When the light in the laboratory was turned off, the scientist-observer managed to unnoticeably unscrew some sensor there, and as a result the control device showed just nonsense. But nevertheless, from a pioneer scientist's point of view, the experiment went positively and proved once again the existence of these very rays.

 
Андрей:

Yusuf, answer me too! No possibility to test, when working in test mode the indicator gives such a picture after 100 bars and until the end of testing:


Andrey, the indicator is not ready for testing on the tester yet due to unauthorised overrides and overflows, we are looking for these reasons now. For now, you can install it on the euro/dollar, TF M1 period of 1000. For other instruments, you need to experiment to determine the TF and the period, until it is attached to the Expert Advisor to determine the optimal period for each instrument. For now, you need to believe the indicator readings on the 0th and nearby historical bars. In your case, if you notice during testing, suddenly, after 100 bars, the indicator reaches very high values and for this reason, the former and future "normal" values turn into straight lines, because of the insignificance of the values in comparison with the bursts. The reason for this is possible overflow of some buffer used for operational summation on each tick. This bug should go away soon.
 
Dmitry Fedoseev:

You know, back in the day there was an interesting article about scientific research in Tekhnika Molodezhdy magazine. Some scientist announced the discovery of some rays and decided to show the experiment to some other scientist. When the light in the laboratory was turned off, the scientist-observer managed to unnoticeably unscrew some sensor there, and as a result the control device showed just nonsense. But nevertheless, from the point of view of the pioneering scientist, the experiment was positive and once again proved the existence of these very rays.

Dimitri, did you have time to turn off the market?
 
Yousufkhodja Sultonov:
Andrey, the indicator is not ready for testing on the tester yet due to unauthorised overrides and overflows, we are looking for these reasons now. For now, you can install it on euro/dollar, TF M1 period of 1000. For other instruments, you need to experiment to determine the TF and the period, until it is attached to the Expert Advisor to determine the optimal period for each instrument. For now, you have to believe the indicator readings on the 0th and nearby history bars. In your case, if you notice during testing, suddenly, after 100 bars, the indicator reaches very high values and for this reason, the former and future "normal" values turn into straight lines, because of the insignificance of the values in comparison with the bursts. The reason for this is possible overflow of some buffer used for operational summation on each tick. This bug should go away soon.

Ok thanks!

 

I have a question.

From the first post, it is clear how the indicator values are calculated. But it is one line. Why do most of the figures from the second page and further show two lines? What is the second line?

 
Yousufkhodja Sultonov:
Dimitri, did you have time to turn off the market?

No, it's more likely that you haven't turned it on.

Well, for example, the first forecast - this one? It's not a fact that it was, it was drawn on history, and the indicator is a re-runner.

Reason: