Download MetaTrader 5

Error Backtesting : no data restored

To add comments, please log in or register
Interested in MQL5 features? Find them in documentation!
Marcel2010
82
Marcel2010 2011.08.07 11:28 

I am trying to back test a few EA's but get every time the following error message and can't find a proper answer so far on this issue. The error message is as follows:

11:15:20 EA name: optimization started
11:15:20 EA name: no data restored
11:15:20 EA name: optimization stopped
11:15:20 There were 1 passes done during optimization.

I expected a lot more results because I chosen to back test on at least 2 different variables.
Any one any idea what's going wrong and how I can solve this issue?

Simon Gniadkowski
Moderator
17998
Simon Gniadkowski 2011.08.07 11:43  

Have you checked that you have history data for the pair that you want to test for the period that you want to test ?

I did a search on "no data restored" didn't come up with much . . .

zzuegg
1505
zzuegg 2011.08.07 14:09  
Have you set the range of the two variables?
Marcel2010
82
Marcel2010 2011.08.07 15:23  
RaptorUK:

Have you checked that you have history data for the pair that you want to test for the period that you want to test ?

I did a search on "no data restored" didn't come up with much . . .


Thanks RaptorUK for your answer. Yes, I have historical tick data from 2007 (file is bigger than 2 Gb, I have run a patch to remove the volume limit in order to make it possible for MT4 to use FXT files which are bigger than 2 Gb).

When I just do a back test without optimization and in visual mode than it's no problem (but than I can't optimize and that's what I want to do)

Marcel2010
82
Marcel2010 2011.08.07 15:25  
zzuegg:
Have you set the range of the two variables?

Hi Zuegg, yes, I have checked the boxes in front of the variables in the input screen when I start optimization and chosen a reasonable range (start - step -stop).
Simon Gniadkowski
Moderator
17998
Simon Gniadkowski 2011.08.07 15:38  
Marcel2010:

Thanks RaptorUK for your answer. Yes, I have historical tick data from 2007 (file is bigger than 2 Gb, I have run a patch to remove the volume limit in order to make it possible for MT4 to use FXT files which are bigger than 2 Gb).

When I just do a back test without optimization and in visual mode than it's no problem (but than I can't optimize and that's what I want to do)

Ah you are using Dukascopy tick data . . . I've just finished saving 5 pairs worth of data as CSV and I have another 2 pairs downloading . . . not tried it in MT4 yet.
Simon Gniadkowski
Moderator
17998
Simon Gniadkowski 2011.08.07 15:45  
Did you make sure you deleted all the old data/history files for the pair you are testing with ?
Marcel2010
82
Marcel2010 2011.08.07 15:56  
RaptorUK:
Ah you are using Dukascopy tick data . . . I've just finished saving 5 pairs worth of data as CSV and I have another 2 pairs downloading . . . not tried it in MT4 yet.

Yes, I have used the Dukascopy tick data already many times and successfully but now I get this message. I am going to reinstall them again and see what happens.
zzuegg
1505
zzuegg 2011.08.07 16:05  
Duka datafeed works good for me
Marcel2010
82
Marcel2010 2011.08.07 21:55  
I have reinstalled the Dukascopy data but still get the same errror message.
David
13
David 2012.01.04 07:06  

I'm getting the same error using tick data while back testing. I also have tick data from 2007 to present. It seems that the optimization will work without the error if I run with a date range spanning the entire tick data (i.e., from 2007 to present). Otherwise, if I use a smaller date range, or in other words, a subset of the tick data, I get the "no data restored" error and the optimization ends after the first pass. This makes it tough to optimize for smaller ranges.

I suppose I could build individual .fxt files for different date ranges, but that seems like a lot of work, when it should "just work."

Marcel2010, did you have any luck in overcoming this error? Has anyone else experienced this error and have an idea how to work around or overcome?

12
To add comments, please log in or register