Testing 'CopyTicks' - page 32

[[Deleted]]

coderex:

somewhere on the forum met an explanation by Renat about the amount of data received CopyTicks, please, who knows this information here

Added: found where it was discussed

Who's testing CopyTicks, has anything changed since that discussion, in the documentation is only an old implementation of this function

If you're talking about the cache, then personally, at this stage, it's not particularly interesting to me. It will be interesting when CopyTicks() will work without errors.

[coderex]

Alexey Kozitsyn:
If you mean the cache, then, personally, I'm not particularly interested in it at this stage. It will be interesting when CopyTicks() will work without errors.

And what errors have you found in CopyTicks now? Not out of idle curiosity I ask you, but to systematize them and track the correction.

[[Deleted]]

coderex:
And what errors have you found in CopyTicks now? Not out of idle curiosity, but to systematize them and keep track of fixes.

https://www.mql5.com/ru/forum/42122/page25#comment_2928386

Apart from error 1 - that was my fault.

[coderex]

Alexey Kozitsyn:

https://www.mql5.com/ru/forum/42122/page25#comment_2928386

Apart from error 1 - that was my fault.

If I understand correctly, it makes sense to check the outermost ticks where they go, it's likely to be in the neighbouring bar, try sending the wrong calculation to the file, showing the ticks (their data) that are in the bars.

[[Deleted]]

coderex:

If I understand correctly, it makes sense to check the outermost ticks where they go, it is likely to be in the adjacent bar, try to dump the wrong calculation into the file, showing the ticks (their data) that are in the bars.

No, this control is present. Moreover, without it it is impossible to accurately pack the ticks into a candle. Yesterday we discussed with fxsaber that the same error occurs on one and the same candlestick, on one of two BCS servers.

And, if the extreme ticks hit the adjacent bar, then the control would not pass the volume of the adjacent bar either. And these errors do not go together.

[coderex]

Alexey Kozitsyn:

No, this control is present. Moreover, without it it is impossible to accurately pack ticks into a candlestick. Yesterday we discussed with fxsaber that the same error happens on the same candle, on one of two BCS servers.

And, if the extreme ticks hit the adjacent bar, then the control would not pass the volume of the adjacent bar either. And these errors do not go together.

i.e. it is the curved volume that comes in?

[[Deleted]]

coderex:
i.e. it is the volume curve that comes in?

Sometimes a curved volume comes as a "reference" from the Volume[] array. Sometimes the volume is collected from the ticks of the curve.

[coderex]

Alexey Kozitsyn:
Sometimes a curved volume comes as a "reference" from the Volume[] array. Sometimes the volume collected from the ticks is curved.

So maybe just the history is crooked? Add more data from the exchange as a reference.

[[Deleted]]

coderex:
So maybe the story is just crooked? Add more data from the stock exchange as a benchmark.

How do you add? Please explain.

[Alexey Volchanskiy]

Alexey Kozitsyn:

What kind of testing can we talk about when there are still B/S deals on the MQ demo?

We'll have to wait for the official update...

I'm out of date, can you tell me what B/S deals are?