Errors, bugs, questions - page 2607

[Slava]

Slava:

For historical data, open, high, low, close prices of corresponding bars are taken to recalculate open, high, low, close synthetic bar

There will be a correction in the next build.

If a Ask(EURUSD) function is used, then open+spread, high+spread, low+spread, close+spread values of this symbol (EURUSD in this example) will be used when constructing historical synthetic data

[fxsaber]

Slava:

If the ask(EURUSD) function is used, then the open+spread, high+spread, low+spread, close+spread values of this symbol (EURUSD in this example) will be used when constructing historical synthetic data

This will still be very far from accurate. For example, low_ask_EURUSD is always much larger than low_bid_EURUSD + minSpread. I.e. the synthetic one will obtain much better askskis than the real one (based onthe tick history).

Perhaps, it makes sense to offer to calculate a part of the history precisely - through ticks.

[Lyuk]

Slava:

For the historical data the open, high, low, close prices of the corresponding bars are taken to recalculate the open, high, low, close synthetic bar

This is a clear error, these are bid prices. Everything described here https://www.metaquotes.net/ru/metatrader5/news about the dollar index, etc. is incorrectly based on historical data. I wrote more about price history here https://www.mql5.com/ru/forum/327330.

"then the open+spread, high+spread, low+spread, close+spread values of this symbol will be used to construct the historical synthetic data"

This is also wrong, I agree with the above comment.

[Slava]

fxsaber:

This will still be very far from accurate. For example, low_ask_EURUSD is always much higher than low_bid_EURUSD + minSpread. I.e. the synthetics will have much better askskis than in reality (on the tick history).

Perhaps, it makes sense to offer to calculate a part of the history precisely - through ticks.

Yes, it makes sense to count through ticks. But it is unclear when it will be implemented

[Slava]

Lyuk:

This is a clear error, these are bid prices. Everything described here https://www.metaquotes.net/ru/metatrader5/news about the dollar index, etc. is incorrectly based on historical data. I wrote more about price history here https://www.mql5.com/ru/forum/327330.

"then the open+spread, high+spread, low+spread, close+spread values of this symbol will be used when constructing the historical synthetic data"

This is also wrong, I agree with the above comment.

Wrong. But somewhat closer to correct than the previous one.

At this stage, we cannot offer a more correct variant than this

[Andrey Khatimlianskii]

I understand wanting it "out of the box", but as it is, anyone can build such an index in a castaway tool from any data, including ticks.

[Stanislav Korotky]

How do I know that the optimisation has been interrupted by the user (or for some other reason) and has not finished? It seems that there is no way to do it now? The OnTesterDeinit handler should accept the reason parameter similar to OnDeinit (adding the appropriate code/number).

[fxsaber]

Stanislav Korotky:
How do you know if an optimisation has been interrupted by a user (or for some other reason) and not completed? There seems to be no way to do it now, right? The OnTesterDeinit handler should accept the reason parameter similar to OnDeinit (adding the appropriate code/number).

A complete overkill can be recognised.

[Stanislav Korotky]

fxsaber:

It is possible to find out the complete overkill.

I would like to know the general case, including genetics.

[[Deleted]]

I noticed that MQLInfoString(MQL_PROGRAM_NAME) returns different values in MT5 terminal and MT5 tester.

In the client terminal it returns MyIndicator, while in the Strategy Tester it returns MySubFolder\MyIndicator.ex5

Is this a bug or a feature?