Reserved to sub and read later
This is very useful Clark, thanks for sharing! Esp the data download and convert part.
For testing I’m still willing to use some programming language other than MT4
Thanx for sharing…subscribed! 
Always wondered how to do this…
Many Thanks ClarkFX, I will use this to try.
Regards !
Thank you Clark! This is indeed a good tutorial that I’m going to follow for sure! I need to backtest and optimize two strats that i couldn’t do with my broker’s data!
Thank you for this very informative tutorial. Great job!
Do you have any advice on the analysis itself? What are the main areas we should look at when analyzing an EA?
I am a beginner, but very much willing to learn how to develop reliable EAs…(for myself, of course). I am working on 1 EA now (but it is in early stage) and I will use this technique described by you to tune it, eventually (I hope I will be able to end the development process sometime… ).
Thanks again.
Glad some people are finding it useful. It is a long read. I didn’t expect much people to go through it all.
I have this working and backtesting the MACD ea
ClarkFX, may I ask you how to get 5 digits data, if possible?
Thanks!
Great! I’m glad it’s working out.
Dukascopy data source is in points not pips (0.1 pip increments). So it will already be in 5, or 3 for Yen crosses, digits. Just off the top of my head, the CSV files attempt to save some memory by making quotes that are whole pips only 4 or 2 digits. So 1.39510 will be 1.3951, but 1.67897 will still be 1.67897.
If I am off I will take a look after the long weekend.
Hello,
In terms of analysis, there are various ways you could approach it. Generally, it will come with experience. Every trading system is unique, so I like to think that each system should be developed differently.
Typically, drawdown, return ratios are something to keep in mind. As well, testing in-sample versus out-of-sample data. I would also take a look at walk forward testing. These generally test for reliability of the system.
Again, there are much more techniques, more than a couple bookshelves of text can cover. For instance, something I occasionally look at is the z-score of the algorithm just to optimize money management strategies, more specifically equity curve optimization, of a particular trading system.
Hope this helps.
ClarkFX, may I ask you something?
For some reason the backtester stops before the end of the test. I mean, I setup for 2007/05/01 to 2013/03/28 and it is stopping at about 2010/07
What could be the problem?
Thanks!
EDIT
If i use control points intead of every tick, I get the complete results.
For a strat that uses mainly daily bars, do you think that backtesting with control points on 1 minute timeframe (the EA makes the necesary adjustments to use daily data) is enough to get a good result?
Thanks!
I am guessing that you’re backtest is stopping earlier than it should because you’ve run out of money. I would look at the log to see if that is the case. Obviously if your backtest has reached a level where it can no longer trade, the backtest will stop.
I would use Every Tick for the highest accuracy still.
I checked it an that is not the problem. The equity never drops more than 13%
What does your logs say?
DISREGARD THIS MESSAGE…
I just realized what was going on… the images server you are using… postimg.org for some reason is recognized as unauthorized website by my internet provider… as I am an expat in the UAE, it’s sometimes really frustrating to deal with these things…
I can see your post entirely just only passing through my VPN, witopia.org.
Problem solved…
Hello ClarkFX,
I am new to the forum, also did not post for almost a year.
I was interested by your post, and unfortunately I cannot see the images posted.
Is that only me, and my setup or are they just broken links ?
Regards,
Pete.
Glad you found a way around it. 
Enjoy.
Hi Clark,
I used EA with 4 pairs on demo account… how I can do a backtest with 4 pairs ?.. on MT4 strategy tester only we can use one symbol.
I’m having problems with AUDUSD. I found two bars that have values out of the range. Let say, on April 4 2007, there is one that has it’s high value of 234, and on April 1 2007 the high and close value is 1.6, double than the open.
Don’t know if the problem is on the original data or with the python script.
The best way would be to test each market individually, save the backtest reports and merge them using Strategy Quant’s EA Analyzer.
I looked at the code, there shouldn’t be any issue, if there is, the most likely problem is the Dukascopy data itself had some sort of error during those times. You can take a look in the History Center and if you can find it, you are able to edit them.