|
Time period |
Hoshigakiii
|
Post subject: Time period |
Post rating: 0
|
Posted: Sun 18 Jun, 2017, 18:24
|
|
User rating: 2
Joined: Fri 17 Jun, 2016, 20:22 Posts: 47 Location: Poland,
|
Hi, I. Could you clear up differences between these periods? II. What kind of problem this could be - my strategy (same conditions) gives me different results when i change Begin Date (3. picture). For example from 18/04/2011 to 18/05/2015 - Strategy open possition in other time, even different day. (i use shift 0 and 1 ; indicators candle period daily ; and from image above 1. daily 2. and 3. same as in the picture) Cheers.
Attachments: |
period.png [79.78 KiB]
Downloaded 372 times
|
DISCLAIMER: Dukascopy Bank SA's waiver of responsability - Documents, data or information available on
this webpage may be posted by third parties without Dukascopy Bank SA being obliged to make any control
on their content. Anyone accessing this webpage and downloading or otherwise making use of any document,
data or information found on this webpage shall do it on his/her own risks without any recourse against
Dukascopy Bank SA in relation thereto or for any consequences arising to him/her or any third party from
the use and/or reliance on any document, data or information found on this webpage.
|
|
|
|
|
|
amine_chourou
|
Post subject: Re: Time period |
Post rating: 0
|
Posted: Tue 20 Jun, 2017, 12:31
|
|
Visual JForex expert at Dukascopy | |
User rating: 288
Joined: Thu 18 Apr, 2013, 09:26 Posts: 1496 Location: Switzerland, Geneva
|
Hi, These are 2 different parameters: 1/ Default period : it's your trading period set up in your strategy. 2/ Runtime Period: this is a specific parameter known also as "Interpolation parameters" : this is a data filter applied on historical tester to speed up the testing speed. More about interpolation in : https://www.dukascopy.com/swiss/english ... ion#p79567
|
|
|
|
|
Hoshigakiii
|
Post subject: Re: Time period |
Post rating: 0
|
Posted: Wed 21 Jun, 2017, 08:21
|
|
User rating: 2
Joined: Fri 17 Jun, 2016, 20:22 Posts: 47 Location: Poland,
|
I still don't get it. Could you say it simple? I need to know how it works, because without knowing periods my strategy is useless. for example: Hoshigakiii wrote: Hi, I. Could you clear up differences between these periods? 1. This is period .... and it's used in .... to ... (simply as you can) example: this period sets a time of strategy checking, if it's set on 1min then strategy will check conditions every 1 minute from starting date. It's to make checking faster (please add as many step by step example as you can) 2. 3. Hoshigakiii wrote: II. What kind of problem this could be - my strategy (same conditions) gives me different results when i change Begin Date (3. picture). For example from 18/04/2011 to 18/05/2015 - Strategy open possition in other time, even different day. (i use shift 0 and 1 ; indicators candle period daily ; and from image above 1. daily 2. and 3. same as in the picture) Maybe you could tell me how to set this periods for this conditions, long term strategy. I will be grateful for clearing this.
|
|
|
|
|
amine_chourou
|
Post subject: Re: Time period |
Post rating: 0
|
Posted: Wed 21 Jun, 2017, 09:12
|
|
Visual JForex expert at Dukascopy | |
User rating: 288
Joined: Thu 18 Apr, 2013, 09:26 Posts: 1496 Location: Switzerland, Geneva
|
Alright, I'll try make it as simple as possible: 1- As said, this is your trading default period. when using the block that you show in 2 your strategy is subscribed to 10mn candles. In other words at every previous candle close and new candle start the strategy will "read" conditions that are linked after this block. 3- The interpolation method is a process to speed up historical tester: instead of going through every single tick the strategy will "skip" some prices depending on the method specified. This will make the historical tester processing more fast (example: instead of testing a strategy over last week in 5mn with "All ticks" it will take 2mn with a filtration method applied) When you select "All ticks" this means no price filtration is applied and the HTester will go through every single tick just like a normal LIVE market goes. When you select another method, you apply a price filter: Example: If you select "4 ticks on OHLC" the historical tester takes your default period (10mn in your example) and builds candles based on Open, High, Low and close prices only therefore the engine will ignore any price within this candle except the later 4 prices. Such methods can change the strategy results because they modify the strategy input data (i.e prices): Some strategies implements indicators that uses several ticks within a given candle, when you apply an interpolation method (price filtration) the output of the indicator changes as you modify the price range used to calculate this indicator --> Result is different. More about tick filtration in : https://www.dukascopy.com/wiki/en/forex ... cal-testerCan't make it simpler than that
|
|
|
|
|
Hoshigakiii
|
Post subject: Re: Time period |
Post rating: 0
|
Posted: Wed 21 Jun, 2017, 14:50
|
|
User rating: 2
Joined: Fri 17 Jun, 2016, 20:22 Posts: 47 Location: Poland,
|
So, in my words: 1. Time between conditions check. 2. Similar to 1. but have ability to work between 1. period checks. 3. Price average, that can change results. (speed up process) Something like that? edit1 To make best possible similar conditions to live i should put 1. on 10sec and 3. on ticks? edit2 What if my strategy have still different results, even with 1. 10s ; 2. deleted ; 3. tick? (only change is starting date)
|
|
|
|
|
amine_chourou
|
Post subject: Re: Time period |
Post rating: 0
|
Posted: Thu 22 Jun, 2017, 10:07
|
|
Visual JForex expert at Dukascopy | |
User rating: 288
Joined: Thu 18 Apr, 2013, 09:26 Posts: 1496 Location: Switzerland, Geneva
|
Hi, Well if you understood these concepts "in your words" that's fine but 1- is a variable 2- is a logical condition that uses default variable and 3 is a parameter that belongs to historical tester. If you want your tests to be similar to LIVE conditions then use always "TICKS" in 3. Quote: What if my strategy have still different results, even with 1. 10s ; 2. deleted ; 3. tick? (only change is starting date) This doesn't make sense to me. Again " 1- is a variable 2- is a logical condition that uses default variable and 3 is a parameter that belongs to historical tester. " Please read more about default parameters in the user guide : Help menu > User Guide
|
|
|
|
|
ForexAlyoum
|
Post subject: Re: Time period |
Post rating: 0
|
Posted: Sat 05 Aug, 2017, 11:53
|
|
User rating: 0
Joined: Tue 15 Jan, 2013, 17:30 Posts: 11 Location: Jordan, irbid
|
Hoshigakiii wrote: So, in my words: 1. Time between conditions check. 2. Similar to 1. but have ability to work between 1. period checks. 3. Price average, that can change results. (speed up process) Something like that? edit1 To make best possible similar conditions to live i should put 1. on 10sec and 3. on ticks? edit2 What if my strategy have still different results, even with 1. 10s ; 2. deleted ; 3. tick? (only change is starting date) i think you need to check your indicator time period , all must be the same as the default time period
|
|
|
|
|
|
Pages: [
1
]
|
|
|
|
|