| Pages:
1
..
39
40
41
42
43
..
98 |
tornado
Junior Member
Posts: 9
Registered: 15-6-2020
Member Is Offline
|
|
Hi Peter,
I installed GSB, but I have only 6-day trial instead of 14-day trial.
Why is the case?
Thanks.
|
|
|
jtango88tw
Newbie
Posts: 1
Registered: 23-3-2020
Member Is Offline
|
|
Hi Peter,
I set positive min. profit criteria using performance filter but still saw systems with negative profit showing in unique-systems.
Any idea how to make the performance filter work as expected?
Thanks!
|
|
|
getty002
Junior Member
Posts: 30
Registered: 10-7-2020
Member Is Offline
|
|
Thanks to Daniel and Randy for your feedback. I continue to believe that regimes can be appropriate, and powerful tool to achieve objectives.
Whether we realize it or not, indicator 3 using data 3 is itself a "regime" filter, it's just more challenging to extract what that regime is within
GSB. And regimes certainly don't have to further complicate the system build, in fact I believe it can simplify it. A simple example of this would
be to have 1 of the 3 indicators dedicated to the 50 day SMA. When trading above the 50 day SMA, my remaining two indicators operate in a momentum
approach. A complementary strategy is then constructed for bear market conditions for the 2 indicators when below the 50 day SMA. This is a dramatic
over-simplification for example purposes, but it directly exposes 2 opposing regimes exemplified through just one of the indicators. The edge is
still present in both regimes and allows construction of a net system which is guaranteed not to trade on top of each other, or in opposing
directions.
Regarding not trading for extended periods, this is perhaps a personal preference. For me, given 2 systems that have the same NP and DD, all else
equal, I choose the higher trade count and higher statistical sample set against historical data, despite a lower Profit Factor. I have greater
confidence of my edge given the higher number of trading opportunities presented. As my trade count becomes fewer and fewer, yes my average trade
will go up, but almost inevitably, I need to be exposed to higher stop losses to reach the increased average trade, for example with swing/trend
algos.
Again, thank you for your inputs. Perhaps one day regimes will be something exposed in GSB. Based on my current understanding, I believe the way
that the indicator operators (multiply, add, etc.) make it a bit challenging to do right now. This is a really powerful aspect of GSB, but also
modifies the way the indicators were originally constructed. If I were to speculate, this is also part of the reason why complementary and
anti-correlated markets don't appear to help performance in GSB, although I certainly don't know for sure. I'd welcome input for others here with
experience trying to use DX, TRIN, VIX, A/D, US, etc. on equity markets with GSB. An algorithm description document (ADD) would help tremendously to
explain the details of the calculation that arrives at the entry signal ;-)
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
Hello All,
Question about the difference between Avg A Fitness adn Avg E. Fitness.
A covers dates from 1900-2015, E covers dates from 1900-2018. A does not re-backtest, E does.
The timeframe on E is longer, yet the fitness is always much lower with the re-backtest.
I guess I don't understand what the re-backtest is doing that is reducing fitness by so much.
Thanks again,
Craig
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Hi Bizgozd, Not sure what macro you are referring to, but i would assume that you build on - 2015, i.e A, then backtest to 2018, meaning that you
would get OOS on 2015 - 2018 or perhaps 2000-2018.. then i assume that you have settings that is not optimal or its just not good times for the
market in question after 2015 to 2018 so your OOS results is less great compared to your IS i.e - 2015... Cheers
Thanks received (1):
+1 bizgozcd at 2020-07-26 14:31:49
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
Quote: Originally posted by Daniel UK1  | | Hi Bizgozd, Not sure what macro you are referring to, but i would assume that you build on - 2015, i.e A, then backtest to 2018, meaning that you
would get OOS on 2015 - 2018 or perhaps 2000-2018.. then i assume that you have settings that is not optimal or its just not good times for the
market in question after 2015 to 2018 so your OOS results is less great compared to your IS i.e - 2015... Cheers |
Thanks Daniel, bad question. I looked more closely at the steps in Macro 3 and I misunderstood exactly what Macro 3 is doing.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by tornado | Hi Peter,
I installed GSB, but I have only 6-day trial instead of 14-day trial.
Why is the case?
Thanks. |
I can only assume it was run at an earlier date
Your login was made June 15 2020.
If you email me teamviewer details, I will fix it
peterzwag@gmail.com
Peter
|
|
|
getty002
Junior Member
Posts: 30
Registered: 10-7-2020
Member Is Offline
|
|
I have NG data that goes back to 2003, but I've set the Global Start Date and Trading Start Date both to January 1, 2007. All goes well until I do a
Walk Forward at which point it goes back to 2003, which I don't want it to include. How do I limit the time range over which a Walk Forward is
performed?
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by getty002 | | I have NG data that goes back to 2003, but I've set the Global Start Date and Trading Start Date both to January 1, 2007. All goes well until I do a
Walk Forward at which point it goes back to 2003, which I don't want it to include. How do I limit the time range over which a Walk Forward is
performed? |
WF will use global dates in GSB, not the other date fields.
In my opinion this should not be the case, and it will change at some stage
From memory NG data pre 2006 was too eratic to trade. I could be wrong as havnt checked
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
| Quote: Originally posted by admin |
WF will use global dates in GSB, not the other date fields.
In my opinion this should not be the case, and it will change at some stage
From memory NG data pre 2006 was too eratic to trade. I could be wrong as havnt checked
|
That brings up another question I keep thinking about. If I change the dates to account for several years of poor liquidity, what changes should I
make to the Macros?
|
|
|
NickW
Junior Member
Posts: 29
Registered: 12-2-2020
Member Is Offline
|
|
When I am doing NG runs, my data files do not contain anything before 2-Oct-2006. NG data before that is not enough and very erratic. You can see if
when you look at 1min bars prior to that date.
If you don't have data in your data file prior to a certain date, then WF or any run will not have any trades before then. That's another way to
avoid data to be used.
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Yes i agree Nick, before 2007 i dont trust the NG data to provide any real gainful info..
|
|
|
meldinman
Junior Member
Posts: 45
Registered: 22-6-2020
Member Is Offline
|
|
I'm trying to add a new contract for KC (coffee) and I'm having trouble with it. Any tips or steps for getting the proper info into gsb contracts
list?
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
| Quote: Originally posted by meldinman | | I'm trying to add a new contract for KC (coffee) and I'm having trouble with it. Any tips or steps for getting the proper info into gsb contracts
list? |
Meldi, what is the issue for you? how does the problem show? one trick you can do is to clone an existing contract and just change values.
Have you tried to scroll all the way down the list and input details on the empty cells?
|
|
|
meldinman
Junior Member
Posts: 45
Registered: 22-6-2020
Member Is Offline
|
|
mainly get confused with the different point values/tick values etc
|
|
|
cotila1
Junior Member
Posts: 78
Registered: 8-5-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Carl | | Quote: Originally posted by getty002 | This is a question unrelated to GSB, but unfortunately it won't let me create a new topic as a newbie. BTW, glad to join your community!
My question is related to the topic of how to identify when a strategy has decayed sufficiently to consider making changes. Does anyone have any TS
indicators or quantitative calculations for statistically evaluating when to make a strategy change? I watched this great video by "Trade-Systech"
which shows a T-statistic method for evaluating performance that's outside the distribution.
https://www.youtube.com/watch?v=635u9rolv4I&t=611s
It looks like he's also a GSB user? If you know who he is, perhaps send me his handle so I could IM him. As a reference, another good read on
statistical analysis of trading system performance is here (a bit old, 2014): https://www.stat.berkeley.edu/~aldous/157/Papers/harvey.pdf
Thanks in advance!
|
Hi getty002,
The T-statistic shows if the sample (i.e. most recent 30 trades) could be part of the population (i.e. all trades sofar without the most recent 30
trades) based on average and standard deviation.
If the sample is an outlier with a small chance of only x% that it could be part of the population, you could consider stop trading this strategy.
More info here, https://en.wikipedia.org/wiki/T-statistic
I know who the guy in the video is on the GSB forum.
He follows the forum every now and then.
So if he feels like it, he can react to these posts.
And an updated version of the 2014 document you mentioned can be found here:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2474755 |
Just noticed this post. Yes correct all the observations from Carl.
The monitoring tool explained in the video and that utilizes the T-test is not the only approach to evaluate when/if to disable a strategy.
For instance another test I use is a simple Montecarlo analysis which is available in PA tool of Peter
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
right click a chart in TS will show you.
If stuck compare to a known one like ES
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
And please let us know how KC goes, its an exotic one, and untamed beast so far 
|
|
|
meldinman
Junior Member
Posts: 45
Registered: 22-6-2020
Member Is Offline
|
|
thanks, didnt realize it was those tradestation values. makes sense now. Will share if I can get anything usefull out of it. its a great product to
trade...
Thanks received (1):
+1 admin at 2020-08-03 18:04:09
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
Question about WF processing. Sometimes when running Macro 1, my computer runs 40 WF's at a time, other times it runs 10. Right now it has 12, and
does not seem to be using local CPU (Sys. CPU under Machine Resources is about 4% during this WF run.) How can I speed this up, so that local CPU
gets utilized more efficiently?
App Settings/Machine Resources/CPU Cores - Local WFs is set to 8 (out of 10 cores on this processor). Do any other settings need to be tweaked to get
this moving?
Thank you.
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
| Quote: Originally posted by bizgozcd | Question about WF processing. Sometimes when running Macro 1, my computer runs 40 WF's at a time, other times it runs 10. Right now it has 12, and
does not seem to be using local CPU (Sys. CPU under Machine Resources is about 4% during this WF run.) How can I speed this up, so that local CPU
gets utilized more efficiently?
App Settings/Machine Resources/CPU Cores - Local WFs is set to 8 (out of 10 cores on this processor). Do any other settings need to be tweaked to get
this moving?
Hi There, i would assume if i understand correctly, that the cloud provides you with sometimes 40 workers including your own, and then sometimes 10
including your own, i would say thats normal depending on fellow GSB users.
Or am i misunderstanding something?.
Thank you. |
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Daniel UK1 | | Quote: Originally posted by bizgozcd | Question about WF processing. Sometimes when running Macro 1, my computer runs 40 WF's at a time, other times it runs 10. Right now it has 12, and
does not seem to be using local CPU (Sys. CPU under Machine Resources is about 4% during this WF run.) How can I speed this up, so that local CPU
gets utilized more efficiently?
App Settings/Machine Resources/CPU Cores - Local WFs is set to 8 (out of 10 cores on this processor). Do any other settings need to be tweaked to get
this moving?
Hi There, i would assume if i understand correctly, that the cloud provides you with sometimes 40 workers including your own, and then sometimes 10
including your own, i would say thats normal depending on fellow GSB users.
Or am i misunderstanding something?.
Thank you. |
|
I like 2 wf per worker, but more is more efficient.
However if you do more, the last ones to finish come much later than the first.
so it can be slower to get all wf done.
I can have 100 to 200 workers, so that 200 to 400 wf.
(but i havnt done wf for a long time as thats not my focus at this instant)
you should have say 8+ workers (on high end machine) so should get 16 wf at a time.
I think you can run more workers for wf, than you do for building systems
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
| Quote: Originally posted by Daniel UK1 | Question about WF processing.
Hi There, i would assume if i understand correctly, that the cloud provides you with sometimes 40 workers including your own, and then sometimes 10
including your own, i would say thats normal depending on fellow GSB users.
Or am i misunderstanding something?.
|
Not sure what was going on, but the fix ended up being a reboot. 
Thanks again for your help.
|
|
|
tornado
Junior Member
Posts: 9
Registered: 15-6-2020
Member Is Offline
|
|
Hello everyone，
GSB is a very good software. Many features are impressive. Thanks Peter.
I am using the new version 58.63, and I feel that the system is built much faster. But I have a problem.
1. I use Manager and 2 Workers to build the system. At first it worked fine, but the system generated by wokers didn't seem to be passed to the
manager. The number of unique-systemys of Manager is stuck, while the number of unique-sytemys of workers keeps increasing.
If the system generated by workers cannot be passed to the manager, can I manually import them to the manager?
2. Sometimes workers will automatically run walk forward while generating the system, which is strange. When I used the manager to do walk froward, I
also started 2 workers, but no workers participated in the walkforward at that time.
Is there something wrong with the way I use it?
Any help will be greatly appreciated.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by tornado | Hello everyone，
GSB is a very good software. Many features are impressive. Thanks Peter.
I am using the new version 58.63, and I feel that the system is built much faster. But I have a problem.
1. I use Manager and 2 Workers to build the system. At first it worked fine, but the system generated by wokers didn't seem to be passed to the
manager. The number of unique-systemys of Manager is stuck, while the number of unique-sytemys of workers keeps increasing.
If the system generated by workers cannot be passed to the manager, can I manually import them to the manager?
2. Sometimes workers will automatically run walk forward while generating the system, which is strange. When I used the manager to do walk froward, I
also started 2 workers, but no workers participated in the walkforward at that time.
Is there something wrong with the way I use it?
Any help will be greatly appreciated. |
my wild guess is you have running
gsb in C:\GSB\Cloud Workers
not
C:\GSB\Workers
DO you have resource manager running? It should have worker higher priority than cloud worker
Send me teamviewer.com details if stuck
|
|
|
| Pages:
1
..
39
40
41
42
43
..
98 |
![[*]](images/davis/default_icon.gif){kind=icon}
