Testing 'CopyTicks' - page 27

[fxsaber]

Alexey Kozitsyn:
The SD's application includes a link to my posts in this thread. I have described the problem in detail here. I even gave some methods for solving it.

You must have missed it. The link?

[[Deleted]]

fxsaber:
Missed it, apparently. The link?

https://www.mql5.com/ru/forum/42122/page25#comment_2928153

and beyond.

[fxsaber]

Alexey Kozitsyn:

https://www.mql5.com/ru/forum/42122/page25#comment_2928153

et seq.

Read. So this is not a CopyTicks problem!

[[Deleted]]

fxsaber:
I read it. So it's not a problem of CopyTicks!

What's the problem?

If you can't process everything to one tick, isn't that a CopyTicks() problem?

[fxsaber]

Alexey Kozitsyn:
What's the problem?

The formation of volume fields of bars.

Honestly, I don't even know if I need to analyse bars when there is a tick history.

I can't imagine that my TS will analyze bar-history in the future. I do not need it when there is tick history!

[[Deleted]]

fxsaber:
The formation of volume fields of bars.

No, not all of them... the errors are floating... ... sometimes they are there, sometimes they aren't. It's unlikely it's because of the volumes the indicator outputs to the volume[] array.

[fxsaber]

Alexey Kozitsyn:
No, not all of them... the errors are floating... ...they are there and then they are not. It's hardly because of volumes that the indicator sends to the volume[] array.

CopyTicks does not supply all of the data? We compared the history to another source - no discrepancies were found.

[[Deleted]]

fxsaber:

The formation of volume fields of bars.

Honestly, I don't even know if I need to analyse bars when there is tick history.

I can't imagine that my TS will analyze bar-history in the future. I don't need it when I have tick history!

If you cannot imagine something - it does not mean it does not exist. For example, calculating the delta on a candlestick (buy-sell difference). Losing even 1 tick with high volume - the readings will already be incorrect!

[[Deleted]]

fxsaber:
The same story was compared with another source - no discrepancies were found.

I found at least three variations on the discrepancy. There are also variations on the story. And what I compared (CopyRealVolume()).- is also probably not working correctly.

[fxsaber]

Alexey Kozitsyn:
I found at least 3 variations of divergence. Also variations on history. And what I compared (CopyRealVolume()).- is also likely to work incorrectly.

That's why you don't even want to look at the bars.

I ran this indicator

Everywhere I looked, Volume == TurnOver_BUY - TurnOver_SELL. Didn't do an automatic check, of course.