GSB Forums - Powered by XMB 1.9.11

I am contemplating making SF ga not available for stock indices. This could be set in the contracts field, whats allowed.
The user can of course change this.
Its the most common mistake I see people making on the cloud workers.
My testing shows significantly better results on stock indices with closeD{bpv} (not normalize)
Comments any one?

I am against to implement this feature to all users with exceptions. The reason is that while closeD{bpv} is the best filter to build let's say for a bunch of systems to our portfolio, as we continue to develop systems which are added to our portfolio, we may want to add systems with different characteristics/correlation/expectancy. Implementing this superfilter will limit the freedom of research, systems produced, and a strong standardization of the output generated. I am in favour of more flexibility not the opposite. For instance, You could implement this advanced feature to those that explicit request this features. So new comers will have time to learn the basic and after learn the basic eventually ask to have access to advanced features.

admin - 7-2-2019 at 02:11 AM

coccigelus, you can still do this, but you are just going to have go go to the contracts field and allow it. Few users will want to do this, and all can do it.
Many users however get SF wrong which is in no ones interest. This option gives the best of both worlds.

coccigelus - 7-2-2019 at 02:52 AM

coccigelus, you can still do this, but you are just going to have go go to the contracts field and allow it. Few users will want to do this, and all can do it.
Many users however get SF wrong which is in no ones interest. This option gives the best of both worlds.

Admin, If You are saying that I would be freedom to keep the current flexibility currently offered from the software by just allowing it through a setting then go ahead. Maybe You could also reintroduce the old SF parameter settings...

THX for quick reply.

amodkarve - 7-2-2019 at 05:21 PM

Peter: One thing I'm finding as I'm looking at the provided system as well as the systems I am generating with your methodology is that the overall time in market is very low (1-3%). This is corresponding to ~4-5% CAGR. That seems pretty low for the effort. Are you seeing the same? Any recommendation on how to get more time in the market while still keeping good out of sample results?

admin - 7-2-2019 at 05:32 PM

Quote: Originally posted by amodkarve

Peter: One thing I'm finding as I'm looking at the provided system as well as the systems I am generating with your methodology is that the overall time in market is very low (1-3%). This is corresponding to ~4-5% CAGR. That seems pretty low for the effort. Are you seeing the same? Any recommendation on how to get more time in the market while still keeping good out of sample results?

That doesnt bother me personally. Some markets like NQ made really good money in the last year, (very high by nq standards) though I find NQ a harder market than ES.
I think new/additional secondary filters, and price patterns are going to help this.
SF closed keeps you out of the market a great deal. Changing fitness to np*at^0.5 will help, but I dont think its a good idea. Experiment to your hearts content using the stats function to prove what works. SF work I hope will start fairly soon. But there has been a endless task to add features users & I all want.
As these are all small, they went before new SF which is a big task

Price Data Series Session

JasonT - 10-2-2019 at 01:40 AM

Hi Peter, there are some great new features in the new version. In the attached picture, the price data series allows you to select different sessions. That's really cool. Thanks for including that.

When you use session templates in TradeStation it removes the candles from the screen (see the chart attached). As a result the indicators you place on the chart calculate differently. If you are using them for a strategy this can provide very different results.

Can you please clarify how the selection of the Price Data session will affect the results GSB produces? Specifically, if you were using the default 24/7 session and change it to, say, 0830-1430 does GSB 'remove or hide' from its calculations the bars/candle price data outside those times (and so affecting the indicators and how they calculate their values in the same way TradeStation does) OR does it just not allow system trade entries outside those times (currently what I understand the 'Trading Periods > Times' filter does)?

Many thanks

Jason

admin - 10-2-2019 at 03:17 PM

Hi Jason,
Its acts the same way as TS/MC do. GSB will hide from the indicators the periods outside of the session time, and of course those same outside periods cant execute trades.
The time filter determines times you can trade, but all the times are used in indicator calcs.

Hemmo - 13-2-2019 at 10:42 PM

'Auto Trading' settings on MultiCharts for GSB NQ Sys1 strategy:

Hi, I am trialling GSB and at the same time, trialling the NQ strategy that Peter has provided to trial users. The hope is to put the NQ strategy into production once I have purchased a license so that it starts paying off the license purchase cost. I can then concentrate on thoroughly learning the GSB product so as to start generating strategies for a mix of instruments within a diversified portfolio.

The code for the strategy “GSBSys1_RAW_v1.1_NQ” has been transposed into my MultiCharts platform and set on Auto Trading. The strategy code is generating trades in back-testing and also during the live market hours which is good, however, the code is not always generating an entry when it should on some days. Why I say that is because when I switch off the Auto Trading, the chart is indicating that a trade entry and exit should have occurred on a particular day, however, in reality, the trade did not actually occur.

From everything that I have read online and in forums is that the correct setting of MultiCharts for Auto Trading is very tricky and that just one obscure setting can have a dramatic impact on the actual correct execution of a trade.

Are there any current MultiCharts users out there who are able to provide me with the settings that they have used to successfully execute automatic trades in real time on MultiCharts?

As an example, I am posting screenshots of my existing Auto Trading settings in MultiCharts v12. It would be really appreciated if any MultiCharts users could do the same for their MC settings. Also, all feedback would be most welcomed. Thank you!

admin - 14-2-2019 at 12:11 AM

There is another very simple tool to stop windows update
The first complex tool is here. https://www.intowindows.com/windows-update-minitool-alternat...
The simple one is here
https://www.sordum.org/9470/windows-update-blocker-v1-1/

Gregorian - 14-2-2019 at 06:03 PM

Quote: Originally posted by Hemmo

Are there any current MultiCharts users out there who are able to provide me with the settings that they have used to successfully execute automatic trades in real time on MultiCharts?

Hemmo:

You have identified an important, but unclear, aspect of using MultiCharts. So far I have had best results with the settings you list, with one exception: On Recalculate on Broker Events: Check only Order Rejected, leave the other two unchecked.

I'm still not sure if better results are had with Async vs. Sync mode. I've tried both, and results are different, but it's inconclusive yet which is better. Clearly Sync is easier to manage.

It would be interesting to hear others' experiences, too.

jptann - 15-2-2019 at 12:11 PM

A number of years ago when I had my professional CTA I used MC in real time trading with IB and never had any issues with synch. I think it is more an issue with what broker you use, rather than a MC issue. I was trading 200 contracts of ES at a time (reversing 400) all at open of a bar (market trades) and never had any undue slippage (averaged 1/2 point over 3 years with high volume or $6.25 for slippage + commissions). IB was rock solid.

Now for some brokers like dorman I'm not sure. I did trade there but can't remember. Synch is still the easiest to manage.

Some of the new lower margin brokers might be different for day trades. I will be investigating that shortly.

I remember seeing a script in MC that allowed you to trade any outside broker, but I don't remember where I say it.

jptann - 15-2-2019 at 12:15 PM

One final point. I see this all the time. People will send me results from their new "golden goose" script and I always find they don't either include commission and slippage or no slippage. It is a fools errand to develop systems that use limit trades since there is no assurance that you will get all the trades in real time at the limit price. Only using market prices with slippage can you be assured that your development will represent what will happen under real trading. Just a pointer for the new people to our game

admin - 15-2-2019 at 03:52 PM

Quote: Originally posted by jptann

A number of years ago when I had my professional CTA I used MC in real time trading with IB and never had any issues with synch. I think it is more an issue with what broker you use, rather than a MC issue. I was trading 200 contracts of ES at a time (reversing 400) all at open of a bar (market trades) and never had any undue slippage (averaged 1/2 point over 3 years with high volume or $6.25 for slippage + commissions). IB was rock solid.

Now for some brokers like dorman I'm not sure. I did trade there but can't remember. Synch is still the easiest to manage.

Some of the new lower margin brokers might be different for day trades. I will be investigating that shortly.

I remember seeing a script in MC that allowed you to trade any outside broker, but I don't remember where I say it.

I suspect market slippage is now higher. Had discussions with a broker over this. It might be due to market / order manipulation but im not at all sure. What order method did you use. ie market orders, some sort of smart order, stop entry etc

GSB Not running

kelsotrader - 19-2-2019 at 09:36 PM

Hi Peter. I have not used GSB for several month and got 'Old code' notification on trying to use it. I downloaded and replaced the manager and worker exe files and GSB.pdb. with build 50.72 Now nothing is starting. Have tried build 51.71 still not joy.
So what has happened in the last 4 months ? I cannot find the solution. I don't think it is window security as I am starting in Admin mode. But that's a possibility.
Please advise.

admin - 19-2-2019 at 09:40 PM

Quote: Originally posted by kelsotrader

Hi Peter. I have not used GSB for several month and got 'Old code' notification on trying to use it. I downloaded and replaced the manager and worker exe files and GSB.pdb. with build 50.72 Now nothing is starting. Have tried build 51.71 still not joy.
So what has happened in the last 4 months ? I cannot find the solution. I don't think it is window security as I am starting in Admin mode. But that's a possibility.
Please advise.

Likely the DLL' have changed over this period.
Best to all files update except gsb_ipp3.xml
If stuck send me teamviewer details.
Updates now automatic through resource monitor.

NQ settings required for CFD instrument on GSB Contracts List

Hemmo - 22-2-2019 at 03:49 AM

Hi, are there any GSB users that are using CFDs to trade?

Can anyone advise me on the GSB settings required for a CFD version of NQ in the GSB Contracts List?

Ticks?

Pnt Val?

Digits?

Your help would be appreciated.

admin - 22-2-2019 at 04:43 AM

Quote: Originally posted by Hemmo

Hi, are there any GSB users that are using CFDs to trade?

Can anyone advise me on the GSB settings required for a CFD version of NQ in the GSB Contracts List?

Ticks?

Pnt Val?

Digits?
Im very interested how trading CFD will work compared to futures. If you have interactivebrokers TWS, it can give you the contract info.
I couldnt find the ticker.

Your help would be appreciated.

admin - 23-2-2019 at 01:01 AM

Quote: Originally posted by Hemmo

Hi, are there any GSB users that are using CFDs to trade?

Can anyone advise me on the GSB settings required for a CFD version of NQ in the GSB Contracts List?

Ticks?

Pnt Val?

Digits?
Seeing is not much reply to this, best contact your broker.
I would just use standard nq settings in GSB. Suspect the NQ cfd has less leverage.

Your help would be appreciated.

Few questions about GSB

Stourac - 14-3-2019 at 05:07 PM

Hi Peter,

I have few questions for you:

1- I'm following your website for a while and noticed in the performance report of GSBm2.1ES that the net profit during the year 2018 is different than the net profit of the same year you posted a month or two months ago. Does it mean that in the meanwhile you are improving/updating the ts?

2- What does GUI means? (sorry, I'm a beginner...)

3- Somewhere in the forum I've read that GSB architecture allows to build better systems for ES, crude oil, natural gas. Have you built system on forex with GSB? Are there other markets which you think fit well for GSB system's architecture?

Thanks a lot for your reply and congratulations for such a great tool as GSB.

admin - 14-3-2019 at 05:32 PM

Quote: Originally posted by Stourac

Hi Peter,

I have few questions for you:

1- I'm following your website for a while and noticed in the performance report of GSBm2.1ES that the net profit during the year 2018 is different than the net profit of the same year you posted a month or two months ago. Does it mean that in the meanwhile you are improving/updating the ts?

2- What does GUI means? (sorry, I'm a beginner...)

3- Somewhere in the forum I've read that GSB architecture allows to build better systems for ES, crude oil, natural gas. Have you built system on forex with GSB? Are there other markets which you think fit well for GSB system's architecture?

Thanks a lot for your reply and congratulations for such a great tool as GSB.

Good observation, and good timing to discuss.
There has been an additional exit thats now included in GSB added on gsbsys2.1es. Basically if the secondary filter and or original exit no longer says go long, we exit a long.
Inverse for shorts. This system was the first made on multi time frames. (via a crude proof of concept method) It was never optimized since 10.2008 from memory. I re-optimized it this week, but the results were only mildly better. So Im not decided if I will update the code. I intend to update the web site comments on this, but there are much more pressing issues competing for my time.

2) GUI = graphical user interface.

3) Ive done little with forex, but dont think GSB is strong in this area. Thats likely going to change with GSB 2.0 I hoped the alpha release would be out (for GSB purchasers), but this has been delayed due to unexpected enhancements and bug fixes in GSB 1.051.x

GSB users have done vxx (symbol changed now), fang stocks, ie apple, sqqq, cl,ng,ho,rb, soybeans, nasdaq and lots more. It takes me some time to learn a market and some systems can go well, even though market validation doesnt work so well. I think its much more risky to build a system when market validation doesnt work well.
see https://trademaid.info/video.htm

Great questions to ask, and good place to ask it. Many new users post on new threads, but things are less cluttered if posts go in the right place/

Performance Filter - Latest Days

JasonT - 15-3-2019 at 07:30 PM

Hi Peter,

I want to apply performance filters to most recent in sample data, and see what the out of sample results are.

Can I achieve this by setting Performance Filter - Latest Days and a small Validation %, OR using Trading Periods > Dates and Dates Mode method OR another method?

Thanks,

Jason

admin - 15-3-2019 at 07:36 PM

Quote: Originally posted by JasonT

Hi Peter,

I want to apply performance filters to most recent in sample data, and see what the out of sample results are.

Can I achieve this by setting Performance Filter - Latest Days and a small Validation %, OR using Trading Periods > Dates and Dates Mode method OR another method?

Thanks,

Jason

Hi Jason, there is latest days metrics you can adjust the time period.

Also you can change the dates of the contract file. This is buggy and being worked on now. If you change the end date ie 2018 to 2019, change the first date by a day. Then click over ride original settings. This too has a bug that give exception errors. A work around is to set oos to one cpu core. (under app settings) You then may choose to change dates under trading period, and make dates notrade. If needed I will get 51.30 to use. Some of this is improved but not all of it fixed.

JasonT - 15-3-2019 at 09:14 PM

Hi Jason, there is latest days metrics you can adjust the time period.

Thanks Peter.

Does the latest days apply to the Validation % or only the Training/Test %?

For example if I tested 20 years of data and selected 95% training and 5% validation, and set latest days > # of days to 90, would any portion of the latest 90 days be in the validation data or all in the training data?

Thanks,

Jason.

admin - 15-3-2019 at 09:44 PM

Hi Jason,
Im going to ask the programmer about this. Ive never used the feature as I consider it problematic. If you took the best systems of say the last year, and traded them - I think out of sample results will be worse than the systems you did not select.

JasonT - 15-3-2019 at 10:49 PM

Hi Jason,
Im going to ask the programmer about this. Ive never used the feature as I consider it problematic. If you took the best systems of say the last year, and traded them - I think out of sample results will be worse than the systems you did not select.

Thanks,

it also appears that when I set the trading periods > dates to a date earlier than the last date in the data series I am using that GSB produces no systems.

For example the CL30.HO data that has come with one of the GSB versions has an end date of 8 June 2018. When I set the Trading Periods > Dates > To {a date like 4 May 2018} I get no systems being produced. When I select a date on or after 8 June I get plenty of systems being produced..

admin - 15-3-2019 at 11:01 PM

HI Jason
I will give you 51.30. If that doesnt fix it email me teamviewer details. These are the current issues the programmer is working on now

JasonT - 16-3-2019 at 12:23 AM

HI Jason
I will give you 51.30. If that doesnt fix it email me teamviewer details. These are the current issues the programmer is working on now

Looks like that's fixed it thanks Peter.

admin - 16-3-2019 at 02:14 AM

Quote: Originally posted by JasonT

Hi Jason, there is latest days metrics you can adjust the time period.

Jason
Programer says
"Yes, latest days apply to the Validation % (as well as training and validation)

Latest days are simply the latest N days of the data, regardless of the training, test, and validation %.

So, in your example of 20 year, 95% training, and 5% validation, latest days of 90 will fall completely inside the validation period (which is 365 days)."

JasonT - 16-3-2019 at 08:04 PM

Programer says
"Yes, latest days apply to the Validation % (as well as training and validation)

Latest days are simply the latest N days of the data, regardless of the training, test, and validation %.

So, in your example of 20 year, 95% training, and 5% validation, latest days of 90 will fall completely inside the validation period (which is 365 days)."

Ok thanks Peter

CandyMachine - 29-3-2019 at 10:18 PM

Hi Peter,

I tried downloading the trial again because the first time I downloaded it, I had no clue what I was doing and I don't think I ever even tried adding my own price data and didn't play around with it nearly enough. Now that I've done some learning on the basics premises, I wanted to give it shot again. I was able to download the new version but I'm not seeing any options for Built In Indicators under Strategy. The tab is just non-existent.

Also, I was wondering why you don't have training options for max profit factor etc.? It's my understanding that aiming for a somewhat modest profit factor etc will yield strategies that are more likely to work in real-time.

Thank you

admin - 29-3-2019 at 10:22 PM

Quote: Originally posted by CandyMachine

Hi Peter,

I tried downloading the trial again because the first time I downloaded it, I had no clue what I was doing and I don't think I ever even tried adding my own price data and didn't play around with it nearly enough. Now that I've done some learning on the basics premises, I wanted to give it shot again. I was able to download the new version but I'm not seeing any options for Built In Indicators under Strategy. The tab is just non-existent.

Also, I was wondering why you don't have training options for max profit factor etc.? It's my understanding that aiming for a somewhat modest profit factor etc will yield strategies that are more likely to work in real-time.

Thank you

1) This might be disabled in the trial mode. I can get you a fixed if needed.
Check also tools advanced mode is on.

Under fitness you can choose netproft * profit factor, instead of np* average trade. They will give almost identical results.

A warning, we havnt tested the extensively - but felt the systems with the highest profit factors will degrade more than average. Part of that is high profit factor is normally lower number of trades. Lower number of trades is statistically less robust.

admin - 29-3-2019 at 10:26 PM

Candymachine. The options are here. There is a possibility that trial mode doesnt have them. The logic is trial users have enough things to do and tweak - which is problimatic when many dont know what they are doing. Hence some featurs are hidden till the user buys the product.

CandyMachine - 30-3-2019 at 01:37 PM

Awesome, thank you. The Advanced setting fixed the lack of indicator options. Not sure if I came across correctly or not but I was speaking to the fact that you might not want the profit factor to be too high. I suppose you can just take the ones that seem more realistic from the list after the build... In one of your videos you talk about Adaptrade creating overly complex strategies compared to your system. Adaptrade has a setting for lowering complexity, but do you not think that puts the generated strategies on par with yours? Does your system create better strategies in most cases? What is it about the brains of your software that would allow it to accomplish this? In all reality there's a good chance I end up buying both GSB and AB but I'm just trying to understand the strengths and weaknesses of both systems.

petrfin - 31-3-2019 at 09:42 AM

CandyMachine, I can speak from a perspective of a user of both – AB and GSB.
And GSB is really much advanced comparing to AB.
It is really not only about the strategies itself. GSB is actually a full ecosystem you need to generate strategies that have a good chance to perform OOS. You have cloud power, you have walk forward, you have macros soon, you have endless possibilities to validate strategies etc. This really let you generate and validate strategies that are ready to live trade.
The best is that the product is evolving very, very fast. Look on upgrade history AB/GSB. You can see that GSB is really a product developed by very active trader and that is very important to me.
So do not focus on just a strategies itself (which are kind of hard to compare), focus on the whole ecosystem of what is available.

admin - 31-3-2019 at 07:22 PM

Hi Candymachine, petrfin's reply is good. There is a old video on GSB vs Builder
see second one here
https://trademaid.info/video.htm

GSB has developed a great deal since the video was made. The video was made as a very high % of GSB users have bought or tried AB
You also cant see the development of GSB until you purchase and get into the private forum.
here is a glimpse of updates

avatartrader - 5-4-2019 at 07:59 PM

Quote: Originally posted by petrfin

CandyMachine, I can speak from a perspective of a user of both – AB and GSB.
And GSB is really much advanced comparing to AB.
It is really not only about the strategies itself. GSB is actually a full ecosystem you need to generate strategies that have a good chance to perform OOS. You have cloud power, you have walk forward, you have macros soon, you have endless possibilities to validate strategies etc. This really let you generate and validate strategies that are ready to live trade.
The best is that the product is evolving very, very fast. Look on upgrade history AB/GSB. You can see that GSB is really a product developed by very active trader and that is very important to me.
So do not focus on just a strategies itself (which are kind of hard to compare), focus on the whole ecosystem of what is available.

I own both AB and GSB as well and I would also agree.

Bruce - 5-4-2019 at 09:37 PM

Quote: Originally posted by petrfin

I own both AB and GSB as well and I would also agree.

I also have both. Agree with all of the above and with the structure and processes being developed with GSB particularly around OOS standup to the rigour. The other big one for me was the consistency and accuracy when migrating systems from GSB to TS and be able to replicate the results that were built in GSB.

Simultaneous SL and TL

Gregorian - 1-5-2019 at 04:57 PM

Peter:

When both Static Stop Loss and Trailing Stop (fixed) are set to True, does GSB only use one or the other, or will it try using both exits simultaneously? I've never seen both listed under the Parameters tab, only one or the other. I have an application where I know having both simultaneously would help.

admin - 1-5-2019 at 05:18 PM

Quote: Originally posted by Gregorian

Peter:
Good to hear from you Gregorian.

When both Static Stop Loss and Trailing Stop (fixed) are set to True, does GSB only use one or the other, or will it try using both exits simultaneously? I've never seen both listed under the Parameters tab, only one or the other. I have an application where I know having both simultaneously would help.

What I expect is both to be in the code at the same time, else the trailing stop to have its max value set to static stop. Look at the TS/MC code to double check. Its possible that parameters tab doesnt have the fully detail. I dont use trailing stop so not familiar with it.

Test currency pairs

ssp729 - 6-5-2019 at 03:00 AM

Hello everyone. I am a newbie, and i want to demo GSB software.
I want to test GSB software in currency pairs (6B, 6E)... Do i need to have a forex broker account? And how can i get the data?

Bruce - 6-5-2019 at 06:04 PM

Quote: Originally posted by ssp729

Hello everyone. I am a newbie, and i want to demo GSB software.
I want to test GSB software in currency pairs (6B, 6E)... Do i need to have a forex broker account? And how can i get the data?

if you're currently doing pairs trading, are you able to export the data as a CSV/TXT file from that platform? You can IM me and we can discuss further to find a solution

GSB Not generating systems

appengineer - 7-5-2019 at 05:15 PM

Hi,

I am trying to build an ES system with the following settings
SF - CloseLessPrevCloseDBpv
Bar - 30 min
Session - 0830-1500 - Exchange time
Disable SF - CloseLessPrevCloseD and CloseOverPrevCloseD
Operator - +
Validation - 0
Performance Filter
Pearson - 0.9
Min Profit - 1.8
Min Trades - 180

GSB is not generating any systems, any ideas why?

Thank you

admin - 7-5-2019 at 06:27 PM

Hi,

I am trying to build an ES system with the following settings
SF - CloseLessPrevCloseDBpv
Bar - 30 min
Session - 0830-1500 - Exchange time
Disable SF - CloseLessPrevCloseD and CloseOverPrevCloseD
Operator - +
Validation - 0
Performance Filter
Pearson - 0.9
Min Profit - 1.8
Min Trades - 180

GSB is not generating any systems, any ideas why?

Thank you

sf should not be disabled.

appengineer - 7-5-2019 at 06:52 PM

Peter,

Do you mean inside Indicator collection CloseLessPrevCloseD and CloseOverPrevCloseD should be turned ON

Or the SF Mode

See attached

Thanks

appengineer - 7-5-2019 at 06:52 PM

Peter,

Here is the attachment

Thanks

admin - 7-5-2019 at 06:54 PM

Peter,

Here is the attachment

Thanks

What you have done is correct. email me teamviewer.com details and I will look

appengineer - 7-5-2019 at 08:23 PM

Peter,

After additional testing and trial, looks like "Show Systems on Workers GUI" property has to be set to True. If I set it to False no systems are generated.

Thank you

admin - 8-5-2019 at 03:35 PM

Peter,

After additional testing and trial, looks like "Show Systems on Workers GUI" property has to be set to True. If I set it to False no systems are generated.

Thank you

That is supposed to hide systems on the worker, not the manager. I will look into that. Might just remove the option.

fra2019 - 8-5-2019 at 09:37 PM

Anyone got the following error message when trying to automate a GSB system in TradeStation?

admin - 9-5-2019 at 03:47 AM

Anyone got the following error message when trying to automate a GSB system in TradeStation?

your code has > 1 data stream, but chart has only one

ie you might have $spx for data2 etc

appengineer - 9-5-2019 at 05:20 AM

Peter,

GSB Manager became unresponsive after generating about 48000 systems (max set 50000). I looked in my resource monitor and looks like it's still doing something but after hours it's still unresponsive, can I kill the process.

Will it be able to resume where it left?

Thank you

admin - 9-5-2019 at 04:20 PM

Peter,

GSB Manager became unresponsive after generating about 48000 systems (max set 50000). I looked in my resource monitor and looks like it's still doing something but after hours it's still unresponsive, can I kill the process.

Will it be able to resume where it left?

Thank you

What version of GSB are you on?
There was a fix for high ram and cpu a few versions ago.
Have you ran out of ram? How much ram do you have?
There is an option to save systems if certain performance figures are reached, but I have it turned off myself as long term it would use a lot of my dropbox space

admin - 9-5-2019 at 04:55 PM

Hi appengineer
RM is more focused on worker cpu, though it shows total cpu. Best check the ram and cpu useage in windows task manager

appengineer - 9-5-2019 at 06:55 PM

Peter,

My computer ran out of RAM (90% in use), it has about 8GB, GSB Manager is using about 6GB.

I managed to pause the process at about 49,100 systems built.

I will let it run overnight.

I am unable to connect to the server to monitor the workers but I think GSB is processing, as it is writing to disk.

Thank you

appengineer - 9-5-2019 at 06:57 PM

Hi appengineer
RM is more focused on worker cpu, though it shows total cpu. Best check the ram and cpu useage in windows task manager

GSB Manager
RAM is 90%+ used, CPU 5% and disk write/read 100%

admin - 9-5-2019 at 06:57 PM

Peter,

My computer ran out of RAM (90% in use), it has about 8GB, GSB Manager is using about 6GB.

I managed to pause the process at about 49,100 systems built.

I will let it run overnight.

I am unable to connect to the server to monitor the workers but I think GSB is processing, as it is writing to disk.

Thank you

50k systems on 8GB ram is not a good combination.
You can reduce the amount of cores used under app settings, which might use less ram at the cost of speed.
Would be good if you can upgrade the ram or machine

appengineer - 10-5-2019 at 07:45 PM

Peter,

Can GSB leverage GPU?

Thank you

fra2019 - 10-5-2019 at 08:04 PM

When I download historical data from TradeStation, time are based on my computer time i.e. market opens at 21:30 and closes at 04:00 the following day. I understand I need to change the contract table with MOC From 0400 to 0459. Can GSB understand this is following day (so for instance, a session would be Monday 21:30 to Tuesday 04:00) ?

admin - 10-5-2019 at 08:47 PM

When I download historical data from TradeStation, time are based on my computer time i.e. market opens at 21:30 and closes at 04:00 the following day. I understand I need to change the contract table with MOC From 0400 to 0459. Can GSB understand this is following day (so for instance, a session would be Monday 21:30 to Tuesday 04:00) ?

I think this can be done.

There are similar issues for me in Australia, so I keeping my trading computer in central usa time. This fixes other issues like daylight savings. usa and au change at different times. There are also rare cases where closed(1) doesnt work properly in australia time zone

appengineer - 12-5-2019 at 10:24 AM

Hi Peter,

Looking at "Proving the methodology" article, we are using
0% of the data for Test(Optimization settings)

but the Performance Filter for Test is set to
Min Pearson:0.9,
Min Profit Factor:1.5,
Min Trade Count: 100)

Are these settings correct? I was under the impression setting Test % to 0, will result to no test phase but 100% training with nth out of sample.

Also when enabling the best indicators for ES, I don't see any indicators prefixed with GSB_ do I need to load them as custom?
Or by selecting AverageFc that equates to GSB_AverageFc

Thank you

admin - 12-5-2019 at 09:27 PM

Hi Peter,

Looking at "Proving the methodology" article, we are using
0% of the data for Test(Optimization settings)

but the Performance Filter for Test is set to
Min Pearson:0.9,
Min Profit Factor:1.5,
Min Trade Count: 100)

Are these settings correct? I was under the impression setting Test % to 0, will result to no test phase but 100% training with nth out of sample.

Also when enabling the best indicators for ES, I don't see any indicators prefixed with GSB_ do I need to load them as custom?
Or by selecting AverageFc that equates to GSB_AverageFc

Thank you

All settings used are here
https://trademaid.info/gsbhelp/Provingthemethodology.html
Set pearsons to 0.95, and min trade count to 150
Test is ok at 0, as we are using nth and dates to get OOS periods.

There is still a number of things I need to refine and test further.

This is what I am testing to see if it helps.
Using only + operatrs and weights of only 1.
SO no "-" operand, and no weights -1,1,step2

Note also that when you get the top 2000 persons systems, and the take the top 1000 np/dd of the 2000, there is very little over lap between choosing np/dd first vs pearsons.

I dont yet know what is best if any

appengineer - 13-5-2019 at 05:10 PM

Thank you Peter!

Is the idea to follow the same principle(proving the methodology) and apply it to different markets, as we tweak the settings and see what works best?

Thank you

admin - 13-5-2019 at 05:16 PM

Appengineer, absolutely correct. So this also lets you validate what I have done. I have done 5000 WF in the last day, so lots more to investigate.
So far it looks like + operand with no "-" operand and weights all 1 are best.

There is a lot more scope to tweak. IE do we wf with weights 1, or a range.

Proving the methodology

appengineer - 13-5-2019 at 05:52 PM

Appengineer, absolutely correct. So this also lets you validate what I have done. I have done 5000 WF in the last day, so lots more to investigate.
So far it looks like + operand with no "-" operand and weights all 1 are best.

There is a lot more scope to tweak. IE do we wf with weights 1, or a range.

I built about 50,000 ES systems using the settings described in the proving the methodology document but I am getting different results. see attached.

I have observed that the GSB process has stopped but the Manager is using a lot of RAM. The Last Update timestamp keep on updating making me believe there is some process still going on.

Any idea why the degradation result differs significantly?

By the way I used the 16 indicators listed on proving the methodology

admin - 13-5-2019 at 06:12 PM

Appenginer
there is something very wrong. The macro should give stats a to f, which you dont have.
Stats G is something I just added. It is the stats of all 50,000 systems with nth all, dates 2000 to 20150630

If you compare stats A to StatsG, you can see how the 5000 systems I picked compared to the 50,000
(mildly interesting but not important at this stage.
Best you send me teamviewer details.

I suspect you are not running the same macro as me.

appengineer - 13-5-2019 at 06:35 PM

Appenginer
there is something very wrong. The macro should give stats a to f, which you dont have.
Stats G is something I just added. It is the stats of all 50,000 systems with nth all, dates 2000 to 20150630

If you compare stats A to StatsG, you can see how the 5000 systems I picked compared to the 50,000
(mildly interesting but not important at this stage.
Best you send me teamviewer details.

I suspect you are not running the same macro as me.

I think you are right, I didn't use the correct macro and my global end date was also off (2018-02-28 rather than 2015-06-30)

Let me give it one more try if it is still different I will send u a teamviewer

appengineer - 13-5-2019 at 08:02 PM

Hi Peter,

How can you tell when GSB has finished building systems.

I did set the Max number of unique systems to 50,000.

The status says Stopped, but the number of unique systems still keeps on going up - currently at 53,940 unique system built

Thank you,

admin - 13-5-2019 at 08:09 PM

Hi Peter,

How can you tell when GSB has finished building systems.

I did set the Max number of unique systems to 50,000.

The status says Stopped, but the number of unique systems still keeps on going up - currently at 53,940 unique system built

Thank you,

set grace period under app settings to say 60, = 60 seconds.
this ignores all systems coming in after its stopped.
However grace period doesn't work perfectly.

It doesn't cause problems with the stats though, but I put a delay in the macro to over come this small bug.
A warning, don't cancel macros. If you do, no future macros run. Will get this fixed asap

avatartrader - 14-5-2019 at 12:10 PM

This is not so much a support issue as a general question:

Once I have completed a build process, walked forward systems, applied the settings and filtered the systems with astab >= 40, I like to save the systems for later review and analysis, or for future selection for monitoring/trading.

When I save the systems, it appears that the current applied parameters are serialized out with the system correctly (using current WF parameters and OOS = true), but it does not seem to serialize the setting as to whether or not the current WF parameters are applied or a flag that it was walked forward at all.

I realize this is mainly cosmetic as long as it saves and re-applies the right parameters when I re-open the system. Right now, my process is to only save the top walked forward systems anyway, so I just wanted to make sure I am understanding correctly and that what I am saving for later review and use are maintaining all of the prior work that I did to build and walk them forward.

admin - 14-5-2019 at 05:19 PM

This is not so much a support issue as a general question:

Once I have completed a build process, walked forward systems, applied the settings and filtered the systems with astab >= 40, I like to save the systems for later review and analysis, or for future selection for monitoring/trading.

When I save the systems, it appears that the current applied parameters are serialized out with the system correctly (using current WF parameters and OOS = true), but it does not seem to serialize the setting as to whether or not the current WF parameters are applied or a flag that it was walked forward at all.

I realize this is mainly cosmetic as long as it saves and re-applies the right parameters when I re-open the system. Right now, my process is to only save the top walked forward systems anyway, so I just wanted to make sure I am understanding correctly and that what I am saving for later review and use are maintaining all of the prior work that I did to build and walk them forward.

The most important thing is you save the eld, and you also want to save the WF version of the eld. My practice is to save it into TS. I didnt think the WF parameters save, but I havnt checked. The GSB system can be re-created out of the binary hash at the end of the code. But its nice still to save the GSB system too. Does this answer your question?

avatartrader - 14-5-2019 at 07:00 PM

Yes it does. I didn't save the EL code this last time, but I do have a couple of the systems that I had previously ported the EL code for, so I can compare and see what it actually saved since those are using the WF.

What is extremely strange is that neither of the id's of the systems I exported as EL match any of the systems I saved, even though I know for a fact I saved off the same systems.

Even the date portion of the id is wrong. the saved systems all have 20190507, which is when they were built, but the EL code for the couple that I am testing is 20190508.

So, I think you must be right - it looks like the WF EL code is given it's own unique ID different from the system ID, and that is not preserved for sure. Unfortunately, since the ID of the original system and the WF code I do have is different, I can't easily go back and find the system to see if the current parameters after loading are what were the WF parameters at the time.

admin - 14-5-2019 at 07:24 PM

Hi Avatartrader,
I think the ID is made with a date/time stamp in it to avoid duplicate system names. Best save your top systems into ts and or save them with a specific name in GSB.
ie my topESsystem1

avatartrader - 14-5-2019 at 07:32 PM

I just re-ran WF on 1 of the systems that I loaded, and upon completion and setting Use WF = true, the ID within the EL code of the WF versions was different, so that explains that.

In my opinion, that ID should either remain constant, or there should be a variable in the code corresponding to the original system ID to at least enable someone to reference the original system in the event that they need to save and reload for some reason.

I also saved and reloaded that system and did confirm that it does not preserve the WF parameters even when applied. So, that means that if you do a large number of WF and need to reload for any reason without saving the WF ELD's and then reload, then that work is lost.

So, it might be nice to enhance the export process to export not only the system, but also be able to export all of the EL code versions in txt format (including WF if performed) and do so as a sub-folder per system or in zipped format. Since I had a batch of 255 high quality systems, it's not feasible to go and do all of that manually.

Another option is to export the WF settings and data in the serialized system so that the state in which the system was saved can be fully restored when it is loaded.

At minimum, though, it would be very beneficial to at least be able to batch export WF code and have some reference back to the original system ID included.

admin - 14-5-2019 at 08:01 PM

Hi Avatartrader,
You can save your 255 systems in a subfolder of your choice.
I like the idea of saving wf version of systems, and have requested the programmer, to see if this can be done

Is txt of the system needed? If you could reload the 255 systems with the wf parameters, its a simple matter to cut and paste into ts/mc

avatartrader - 14-5-2019 at 08:16 PM

Hi Avatartrader,
You can save your 255 systems in a subfolder of your choice.
I like the idea of saving wf version of systems, and have requested the programmer, to see if this can be done

Is txt of the system needed? If you could reload the 255 systems with the wf parameters, its a simple matter to cut and paste into ts/mc

No, if the current functionality could be enhanced and the system could be saved and preserve the WF parameters/version and other properties, then that would be ideal and GSB could re-generate the required code versions on the fly anyway.

admin - 14-5-2019 at 08:20 PM

Hi Avatartrader,
I will let you know if we can do this, when I hear back from programmer.

engtraderfx - 15-5-2019 at 04:46 AM

I second saving the WF cases, I meant to bring that up myself, would be useful thanks as we lose all the walkforward results if have to close down computer etc.

appengineer - 16-5-2019 at 03:49 PM

Hi Peter,

I am Walking Forward about 1000 systems and it's taking too long, more than 24 hr passed and it's only 25% complete.

Running it on e2680v2 and I chose Genetic Algorithm Multi-threaded

Thanks

admin - 16-5-2019 at 05:00 PM

Hi Peter,

I am Walking Forward about 1000 systems and it's taking too long, more than 24 hr passed and it's only 25% complete.

Running it on e2680v2 and I chose Genetic Algorithm Multi-threaded

Thanks

That doesnt surprise me that it takes that long. I have about 7 servers/pc so can do 1000 in well under a day easily. You can reduce the amount of population & generations if you use 2 indicators. 120x120 to say 80 x 80
Likely if you do 500 wf, the results will be similar. But the more the better as too small sample size will give more variation on the same test.
All you can do is rent (off me )or buy more servers, or wait.
The wf will be 10 to 100 times faster than TS, so you have to put that in perspective. Few people could even imagine its possible to wf 1000 systems over 4 days.

This is one of the reasons I encourage resource manager to be used.
Collectively it gives us all more power the bigger the GSB user base gets.

My own cloud which I donate to users when not used - has been used about 80% of the time in the last week. (much higher than normal) Ive done about 20,000 wf this last week.
If my usage of my servers drops, likely it will help you a bit. Renting a second server for $10 a day also will help you.

admin - 16-5-2019 at 05:26 PM

appenginer
this MIGHT help wf speed. Only good if you have lots of ram.
https://trademaid.info/forum/viewthread.php?tid=92
Increase you cache size, under app settings.

appengineer - 16-5-2019 at 07:50 PM

appenginer
this MIGHT help wf speed. Only good if you have lots of ram.
https://trademaid.info/forum/viewthread.php?tid=92
Increase you cache size, under app settings.

Thank you Peter

Just wanted verify I didn't do something incorrectly.

I have enough memory, the Manager seems to be doing the heavy work, RAM usage is about 170GB, how did you span the WF in multiple servers?
I am planning to replace my PC with a more performant one, it may help share the load.

I will increase the cache size after the WF is complete

admin - 16-5-2019 at 07:55 PM

appenginer
this MIGHT help wf speed. Only good if you have lots of ram.
https://trademaid.info/forum/viewthread.php?tid=92
Increase you cache size, under app settings.

The manager shouldnt be doing the heavy work, it should be the workers.
Are you sending to the workers?
see this to get wf to your cloud and the public cloud
https://trademaid.info/gsbhelp/11GSBFamiliarisationGSBStanda...

appengineer - 16-5-2019 at 09:35 PM

appenginer
this MIGHT help wf speed. Only good if you have lots of ram.
https://trademaid.info/forum/viewthread.php?tid=92
Increase you cache size, under app settings.

The keys are setup correctly, when I built the systems, I saw them appear on the workers.

Both Manager and workers are running on the same server.

It's hard to tell if anything is going on on the workers.

I can see the WF progress in the Manager. No systems on the workers.

admin - 16-5-2019 at 09:36 PM

the workers should be using a fair amount of cpu. send me teamviewer.com details and i will look

admin - 16-5-2019 at 10:59 PM

So, I think you must be right - it looks like the WF EL code is given it's own unique ID different from the system ID, and that is not preserved for sure. Unfortunately, since the ID of the original system and the WF code I do have is different, I can't easily go back and find the system to see if the current parameters after loading are what were the WF parameters at the time.

52.32 has the find system option. Saving WF parameters is in the pipeline

boothy - 17-5-2019 at 12:57 AM

Hi Peter,

I am trying to get caught up with all the new features. It says in the user guide that new macros will be included in latest version of GSB. I have updated to 52.30 but cannot find IndicatorStats.gsbmacro, I can't find it on forum either. Can you please let me know where to find this file?

Thanks.

admin - 17-5-2019 at 01:01 AM

Quote: Originally posted by boothy

Hi Peter,

I am trying to get caught up with all the new features. It says in the user guide that new macros will be included in latest version of GSB. I have updated to 52.30 but cannot find IndicatorStats.gsbmacro, I can't find it on forum either. Can you please let me know where to find this file?

Thanks.

Can you check here
...Data\Settings\Macros

Also go to macros, user macros (right click) macros ..load. and see if the files are there.

Lots of new toys to play with of recent

boothy - 17-5-2019 at 02:24 AM

Quote: Originally posted by boothy

Can you check here
...Data\Settings\Macros

Also go to macros, user macros (right click) macros ..load. and see if the files are there.

Lots of new toys to play with of recent

There was nothing there after updating to 52.30, I found a zip file in the forum which had
alldatesto20150630.gsbmacro
NtahAll.gsbmacro
wf_stats.gsbmacro
and I put them into that file location. But have not been able to find indicatorstats.

admin - 17-5-2019 at 03:06 AM

Here is indicator stats, but its very easy to make the macro yourself if you dont have it.
Other macros are included too.
Inc the one that puts the top 1000 fitness systems into favorite A

Attachment: Login to view the details

avatartrader - 17-5-2019 at 07:24 PM

That's great news - can't wait. Thanks, Peter!

52.32 has the find system option. Saving WF parameters is in the pipeline

boothy - 18-5-2019 at 06:23 PM

Hi Peter,

I have set the parameters as per step 3 in GSB guide for methodology

Use only the addition operand.
Use weights 0.5 to 2 step 0.25 for all weights. Do NOT use subtract or negative weights.

I tried building systems on SI first, which GSB did do but avg trd was really low <$10 for all systems.
Then I tried building on ES and it would not build any systems.

what might cause this?

Thanks.

admin - 19-5-2019 at 04:22 PM

Quote: Originally posted by boothy

Hi Peter,

I have set the parameters as per step 3 in GSB guide for methodology

Use only the addition operand.
Use weights 0.5 to 2 step 0.25 for all weights. Do NOT use subtract or negative weights.

I tried building systems on SI first, which GSB did do but avg trd was really low <$10 for all systems.
Then I tried building on ES and it would not build any systems.

what might cause this?

Thanks.

my guess is you dont have secondary filter on.
slippage and commision set to zero?
send me teamviewer details if stuck.

Look also at thus post
https://trademaid.info/forum/viewthread.php?tid=186
search on "Cant get any systems to appear."

fra2019 - 21-5-2019 at 06:18 PM

was there any change in the display of the results i.e. 60k is now 60 ? If yes, does it impact the input of maximum loss for a stop-loss ? (3k would be 3 ?)?

admin - 21-5-2019 at 06:46 PM

was there any change in the display of the results i.e. 60k is now 60 ? If yes, does it impact the input of maximum loss for a stop-loss ? (3k would be 3 ?)?

can you explain more what you mean? The context will help me be more clear

fra2019 - 21-5-2019 at 06:57 PM

in the tab "Unique systems', profit numbers and drawdowns are numbers like 60 or -3 (for 1 NQ future), while yesterday they were 60k and -3k. Maybe I changed one of the settings by mistake. Will double check tonight

admin - 21-5-2019 at 07:03 PM