History Centre updated - free history of minute quotations since 1999 - page 5

New comment
 
2007.04.14 01:38:20 HistoryBase: not enough memory 'EURUSD1' [2911923 bars]
2007.04.14 01:38:20 Memory handler: cannot allocate 128135876 bytes of memory
2007.04.14 01:35:23 HistoryBase: not enough memory 'EURUSD1' [2911922 bars]14 01:35:23 HistoryBase: not enough memory 'EURUSD1' [2911922 bars]
2007.04.14 01:35:23 Memory handler: cannot allocate 128135832 bytes of memory
2007.04.14 01:33:01 HistoryBase: not enough memory 'EURUSD1' [2911921 bars]
2007.04.14 01:33:01 Memory handler: cannot allocate 128135788 bytes of memory
2007.04.14 01:32:46 HistoryBase: not enough memory 'EURUSD1' [2911920 bars]
2007.04.14 01:32:46 Memory handler: cannot allocate 128135744 bytes of memory
2007.04.14 01:32:44 HistoryBase: not enough memory 'EURUSD1' [2911920 bars]
2007.04.14 01:32:44 Memory handler: cannot allocate 128135744 bytes of memory

window bars 2147483647
build 203
demo account active
 
That's the answer - the terminal lacks the memory in order to upload the quotes at the full depth.

M1 on EUR since 1999 takes 128 Mb and the computer does not have enough memory:

Important: Using deep history requires a lot of computer resources. It is desirable to have at least 1-2Gb of RAM and a powerful processor for comfortable work. We strongly recommend that you monitor the "Tools - Settings - Charts - Maximum bars in window" parameter and set it between 65 000 and 128 000 bars. The "Maximum history bars" parameter can be set high, for example 10 million - it will not seriously affect the terminal's performance.

 
I thought 1.3 gb of memory would be enough
if I put 65,000 bars in the window, test from 1999 and open the chart, will my chart start from 1999 or just 65,000 from the current day?
 

When visually viewing the eurusd minutes, it is striking that the nature of the chart does not resemble what is observed online. The bars are too homogeneous in scope, the behaviour of the volumes is simply not believable. Frankly speaking, I think that the quotes are simply generated by some older timeframe. If so, what timeframe (that is, starting from which one can actually use them)? If not, what is the reason for such an unusual type of charts?

 
The quotes are not generated from the upper timeframes. These are real quotes.

The market is constantly changing. Do not forget that quotes from 1999-200s were made under completely different trading conditions, when spreads were 2-3 times higher and the Instant Execution mode was not used. With the reduction of spreads and the switch to Instant Execution (trading at prices from Market Watch), the requirements to the quality of quotes became very strict. This led to a narrowing of price chatter and a clear reduction in the number of ticks per time unit.
 
Renat писал (а):
The quotes are not generated from the upper timeframes. These are real quotes.

The market is constantly changing. Do not forget that quotes from 1999-200s were made under completely different trading conditions, when spreads were 2-3 times bigger and the Instant Execution mode was not used. With the reduction of spreads and the switch to Instant Execution (trading at prices from Market Watch), the requirements to the quality of quotes became very strict. This led to a narrowing of price chatter and a clear reduction in the number of ticks per time unit.


Looking at 1999-2000 I told myself roughly the same thing. But here is a fresh snippet

But after thinking about it I understood that it doesn't refute you (because this and other problematic fragments are simply absent in the history from Alpari - apparently this is the broker's off-hours). So the market may have changed, but not that much. But it seems that the broker's quotation filtering has radically changed. It is clear that with narrowed spreads the absence of additional filtering would be simply suicide for the broker and of course we cannot reproach him for this. But The problem is that the quotes from History Center cannot be used for testing (except for the generation of very large timeframes). In fact, what is the point of testing a strategy in conditions that did not exist long ago?

 
By the way, here's a thought: couldn't we take a tick story and build the minutes, filtering it by "modern"? I think such a story would be much more useful, even though it didn't really happen :)
 
lna01:


Looking at 1999-2000, I told myself roughly the same thing. But here is a fresh snippet

However, on reflection, I realized that it does not refute you (because in the story from Alpari this and other problematic fragments are simply absent - apparently this is the broker's off-hours). So the market may have changed, but not that much. But it seems that the broker's quotation filtering has radically changed. It is clear that with narrowed spreads the absence of additional filtering would be simply suicide for the broker and of course one cannot blame him for this. But The problem is that the quotes from History Center cannot be used for testing (except for the generation of very large timeframes). In fact, what is the point of testing a strategy in conditions that did not exist long ago?

But this is the Christmas week, when they usually 1) widen the spreads 2) tighten the collateral requirements several times over. Do you still want to trade under such conditions? :)
 
Rosh писал (а):
It's the Christmas week, when they usually 1) widen the spreads and 2) tighten the collateral requirements several times over. Do you still want to trade under such conditions? :)
So this snippet (or snippet from the second of January) only illustrates the problem (and also suggests that some "primary" flow of quotes still exists). And for the earlier data (and not only for 99 and 2000) such quotes are continuous. I'm just interested in as long a filtered history as possible, because the current market will be adequately tested on it.
 
lna01:
Rosh wrote (a):
It's the Christmas week, when they usually 1) widen the spreads and 2) tighten the collateral requirements several times over. Do you still want to trade under such conditions? :)
So this fragment (or the fragment from the second of January) only illustrates the problem (and also says that some "primary" stream of quotes still exists). And for the earlier data (and not just for 99 and 2000) such quotes are continuous. I'm just interested in as long a filtered history as possible, because the current market will be adequately tested on it.
You are following the standard path of beginners, trying to get as low as possible and buried in the tick queue. Roughly speaking - you are engaged in pipsing, constantly depending on a couple of pips of divergence and always in search of an ideal quote.

All this has been discussed many times. Read posts related to this topic and click "similar" at the bottom right of interesting topics. You'll find plenty of similar discussions. For example: Standard misconceptions in trying to trade in noise
123456789101112...21
New comment