GSB Forums - Powered by XMB 1.9.11

Short term progress reports on GSB features

admin - 7-6-2018 at 10:00 PM

This list isn't concrete, but is the intend short term road map forward.
1) Currently we are 80% complete the alpha build of multi time frame & data support.
This means you can build systems on 29, 30,31 minute bars and the system that works with the best parameters on all the times will have the highest metrics.
This should lead to more robust systems.
Also you could use ES,NQ,emd,er,ym etc
or
es data1, $spx data2
emd data1, $idx data2
nq data1, $ndx data2 etc

Later we will use 1 minute input data, so you can optimize bar interval
and just select say 5,15,30,60 min bars that are derived from the 1 minute.

2) csv export of trade lists, not just Portfolio analyst format.

3) Then ability to save averages of all system metrics, to variable storageA
You can then do inverse nth, or invert dates etc, then send to variable storageB
You then get to see the ratios of A vs B. This is for market validation
refer to the youtube video on market validation.
https://www.youtube.com/watch?v=NFC7ego_Y70

4) Smarter stops, exits etc
5) Additional secondary filters.

A lot of work also going into portfolio analysts cloud. Just trialing auto upload of TS trades so results appear on your web site.

admin - 8-6-2018 at 11:16 PM

Multi market validation beta build for GSB purchasers out Monday I hope.
It means you can build on es, $spx and validate on emd$idx, nq $ndx , ym,$dow
I have the build, but no time to test and document it.

admin - 10-6-2018 at 08:06 PM

Out soon 1.0.45.02 / 2018-06-11:
Verification data implemented (as a right click).
Price Data implemented in Tools menu (similar to Contracts).
PerformanceSummary.VerificationScore implemented.
Trades' Export to CSV implemented.
PriceFileInfo implemented to replace PriceFileName.
PriceFileInfos implemented.
Allowed null in DB's ExceptionLog.WorkplaceId.
Minor UI fixes in Admin window, Contracts window, and performance grid.
Minor UI fixes in AdminForm.
ReformatAndRelocateOldFiles() function removed.
OptimizationSettings.ConvertPrimaryAndSecondaryDataToOptimizationData() implemented (to replace primary/secondary data with PriceFileInfos).
Multiple logic updates to the main GSB window.
OverrideOriginalData set to hidden and UpdatePriceData() function deleted.

Cant release this build till the new features are documented. The new features just to hard to use without documentation. Working on this now

Shown is a system built on es data1, with $spx data2.
It's verified on emd $idx, nq $idx, ym $comp, rty $rut etc
Other screen shot is verified on es, $spx 27,28,29,31,32 min bars

admin - 10-6-2018 at 10:52 PM

Here is the docs on the verification data. GSB.exe is in beta section (GSB paid users only)

Attachment: Login to view the details

rws - 11-6-2018 at 10:20 AM

Thank you so much for implementing this usefull feature!!

It would be good if there would be a way of operation where
you can set the additional tickers to have at least xx% profit
of the main ticker in order to have it automatically searching for robust systems.

Quote: Originally posted by admin

Out soon 1.0.45.02 / 2018-06-11:
Verification data implemented (as a right click).
Price Data implemented in Tools menu (similar to Contracts).
PerformanceSummary.VerificationScore implemented.
Trades' Export to CSV implemented.
PriceFileInfo implemented to replace PriceFileName.
PriceFileInfos implemented.
Allowed null in DB's ExceptionLog.WorkplaceId.
Minor UI fixes in Admin window, Contracts window, and performance grid.
Minor UI fixes in AdminForm.
ReformatAndRelocateOldFiles() function removed.
OptimizationSettings.ConvertPrimaryAndSecondaryDataToOptimizationData() implemented (to replace primary/secondary data with PriceFileInfos).
Multiple logic updates to the main GSB window.
OverrideOriginalData set to hidden and UpdatePriceData() function deleted.

Cant release this build till the new features are documented. The new features just to hard to use without documentation. Working on this now

Shown is a system built on es data1, with $spx data2.
It's verified on emd $idx, nq $idx, ym $comp, rty $rut etc
Other screen shot is verified on es, $spx 27,28,29,31,32 min bars

cotila1 - 11-6-2018 at 12:34 PM

Thanks Peter, its a great job. I think the multiple market and 29,30,31 min bars verification is a great step for GSB.
once you release a consolidate build version, would be great you make a new video on this 29,30,31 min bars, and es,emd,ym,nq,rty validation.

thanks

admin - 12-6-2018 at 07:53 PM

Quote: Originally posted by cotila1

Thanks Peter, its a great job. I think the multiple market and 29,30,31 min bars verification is a great step for GSB.
once you release a consolidate build version, would be great you make a new video on this 29,30,31 min bars, and es,emd,ym,nq,rty validation.

thanks

I will do, but there is a lot of features coming in short period, so will wait till the dust settles.

admin - 12-6-2018 at 07:54 PM

Verification score. We now have a count of the number of verification systems that pass the verifcation metrics in app settings.
This should be released in a few days

admin - 12-6-2018 at 07:59 PM

Override original settings. If something on the left opt settings is changed, after systems are built - Then the system results will change when you click on them.
This is wrongly perceived as a bug. To fix this highlight the system(s) and click override original settings.

admin - 13-6-2018 at 12:26 AM

In Build 45.31
We now can do nth ratio calculations fully automated. averageAllSystems(fitness*netprofit*averatetrade*profitfactor)/{nth trade}
averageAllSystems(fitness*netprofit*averatetrade*profitfactor) {Nth dont trade}
run GSB with say nth 1.
If you run the manager (not standalone) wait till new systems top coming through. This can take some time to stop.
Select all system
Right click and change nth day mode from donotTrade to Trade.
Wait till all systems have changed. (popup in windows taskbar will tell you, or scroll through all the systems in GSB to check)
right click and do update statistics.
Ave/previous average. Note in last videos the higher above 1 the ratio is, the worse.
Not we have done the reciprocal of this number. So if our threshold of a market failure was 2, its now 0.5. 1 or above would be perfect score meaning
out of sample performance > in sample performance
This is described in detail in the market validation video https://www.youtube.com/watch?v=NFC7ego_Y70

admin - 13-6-2018 at 05:28 AM

bug, to fix next build
System average should be (fitness*at*np*pf) not fitness+at+np+pf)

Petzy - 13-6-2018 at 05:41 AM

Quote: Originally posted by admin

Override original settings. If something on the left opt settings is changed, after systems are built - Then the system results will change when you click on them.
This is wrongly perceived as a bug. To fix this highlight the system(s) and click override original settings.

I'm not sure I understand this. This is a new behaviour as far as I can remember.
Also I can't see that Overriding the original settings will make any difference. The values still changes

admin - 13-6-2018 at 04:08 PM

Quote: Originally posted by Petzy

Quote: Originally posted by admin

I'm not sure I understand this. This is a new behaviour as far as I can remember.
Also I can't see that Overriding the original settings will make any difference. The values still changes

It might change but thats the settings the system was calculated on. My assumption is settings changed after system building began. I havnt yet seen this issue with the new build, so happy to look via teamviewer when it happens.

rws - 13-6-2018 at 05:37 PM

Good progress

admin - 19-6-2018 at 01:15 AM

Update to documentation for new features of the last week or two.

Attachment: Login to view the details

admin - 25-6-2018 at 12:20 AM

GSB 46.13 should be out later this week. It is very significant in improvements in out of sample results.
If we leave every 2nd day out of sample on ES.30 minutes bars, my tests show a 25% drop in out of sample results.
All systems no matter how bad are included in these results.

Ive got that down to 5.4% drop in out of sample results (best case) compared to in sample.
However my favorite method - that gave the highest fitness (Net profit* average trade) gave 9.4 % drop in out of sample results
but average net profit of $38900 compared to $28,300 out of sample of 30 min bars.
This is research you can objectively test your self, with minutes of human time, and an hour or so CPU time per market.

admin - 27-6-2018 at 03:46 AM

Hoping to upload in a few hours, but have a problem with last part of the video....
Here is a video on some of the features of GSB 47.03
It shows a comparison of a system built on ES, but verified on Dow, Nasdaq, Russell, MidCap S&P400.
It also shows the much faster method to see how much a large group ES30 min system degrades, and then the same test repeated with ES29,30,31 min bars.
Video uploaded but not finished. Hope to have done in the next day or so

video obsolete, see MarketValidation_video2_rev2o.2.mp4 below
Attachment: Login to view the details

admin - 28-6-2018 at 03:23 AM

I started the video from scratch again. One of the many smart GSB users comment that the point value needed for closeD-close offset varies massively for the dow compared to ES, EMD etc, and this might be why so many systems didn't pass on multi market validation. Turns out this was strongly the case. However I got a surprise in results with CloseD(1) /Close compared to closeD(1)-close. The next video shows this.
Video not complete, and about 12 minutes long.
GSB 47.07 is being tested now by some experienced GSB users. It has the features shown in this video.

video obsolete, see https://youtu.be/2R4t9uYzfD4

admin - 29-6-2018 at 02:44 AM

Finished video

It is critical to have good out of sample results, to make money in trading.
Today we are going to have solid evidence showing how much better building systems with multiple time frames is with the same parameter values, and multiple indices with the same parameter values. This data is no derived from the results of one system, but of thousands of systems.
We also cover better alternatives to leaving certain dates for out of sample testing.
The implications to this are far reaching. You can tell what time frame, secondary data source is going to work best on all systems, not just the one you try it on.
Sound logic on 1000's of systems leads to sound logic on one system. Font reverse this concept!

Video now on youtube
https://youtu.be/2R4t9uYzfD4

alvestjo - 1-7-2018 at 04:12 AM

About using closeD(1)-close as a filter. Does this make sense at all if you are using many years of data since the market has traded at very different levels 5-10 years ago. I find it very strange you get better results with this filter. Any ideas why?

Jonas

admin - 1-7-2018 at 05:01 PM

Quote: Originally posted by alvestjo

About using closeD(1)-close as a filter. Does this make sense at all if you are using many years of data since the market has traded at very different levels 5-10 years ago. I find it very strange you get better results with this filter. Any ideas why?

Jonas

I'm not sure why and in once sense I dont care why. I just want the to know what works best, and what doesn't. The first thing I would do is graph range in points of the Dailyclose absvalue(closed(1)-closed(0)) over the years to see the results
One comment is there are some other markets that closeD(1)/close is going to
work better than closeD(1)/close but i have forgotton what ones. I found this by accident. I think for the indices Closed(1)- is best
Run this code on daily bars.
you cloud change
hl = Averagefc(h-l, ATRLength ) ;
to
Averagefc(absvalue(close/open)*100, ATRLength )

admin - 2-7-2018 at 02:59 AM

Coming soon is GUI tweaks, better formatting of the market metrics, and ability to turn of and on sections of the gui you dont need.
I got market degradation down to -1.2%! today on ES.
I will write about in tomorrow in the private form for GSB purchasers

admin - 3-7-2018 at 01:25 AM

The market validation video is now on youtube. Only a few tweaks since the post of MarketValidation_video2_rev2o.2.mp4
Here it is on youtube. https://youtu.be/2R4t9uYzfD4
Crude oil so far out of sample compared to in sample is degradation -35.2%

admin - 17-7-2018 at 01:42 AM

In build 47.41 we can now build systes on say 29,30,31 min bars. Then test the out of sample nth (every second day)
to see the degradation out of sample on 29,30,31 and the individual time frames too.
In the big picture is means we build systems on say 29,30,31 min bars, same parameters for all time frames, but just trade the 30 minute bars. This build will be out soon to in the beta section for paid GSB users, and in a about a week for trial users.

admin - 18-7-2018 at 05:02 AM

For market validation, in build 47.41 we now have long short stats for all time intervals / markets traded.
Useful to tell if a market has a strong long or short bias.

Carl - 18-7-2018 at 07:50 AM

Great! Thanks

admin - 23-8-2018 at 09:50 PM

In build 48.24 Being tested now.
Ability to walk forward multi time frame/market systems to the cloud

Next in the job pipeline, in not exact order
Custom dates for Training, test, validation.
Ability to make larger time frames from small bars. ie make 29,30,31 min bars from 1 minute
Secondary filter. Ability to genetic use Close-Closed,close/CloseD,close-closeBPV (Big point value) in a normalized and non normalized mode.
Custom made metrics in GUI, and additional fitness types. ie sqn and the additional ones in EWFO (Enhanced walk forward optimizer)

cyrus68 - 24-8-2018 at 02:02 AM

GSB has made enormous progress over the past year. The pace of development outshines anything achieved in similar software. Inevitably, there are bugs and annoyances along the way.

The great strength of GSB is the emphasis on good methodology.

zdenekt - 24-8-2018 at 08:54 AM

I still can not see EXITs in the job pipeline :-(
I'm sorry to have to write it, but I'm a little disappointed with the direction of GSB's development over the past few months.
Everything is still spinning around the entering to trade.
We are constantly looking for an ideal entry to trade for a few markets.

In GSB we have thousands of settings for everything that some of us do not even understand, but we have no option to set the type of EXIT ... don´t you think it's a little weird?

I don´t know if there is any major edge in your multitimeframe testing method when the exit from the trade is always at CloseD price. I understand that the results will improve a few percent but it is quite logical when EXIT is on the same price on every timeframe...and programmers still work over and over on one and the same ... better entry on just a three markets.

Yes I can ignore your way of multitimeframe testing, but i´m still waiting for implementation of any different exits.
I believe the implementation of only a few other reasonable EXITs than closeD would not be as problematic as multitimeframe testing and would open a new space of strategies on many more markets than only ES, CL, NG.

Anyway, GSB is a great tool, although it is now going through a more complicated way than it might have been.

admin - 24-8-2018 at 04:36 PM

Quote: Originally posted by zdenekt

I still can not see EXITs in the job pipeline :-(
I'm sorry to have to write it, but I'm a little disappointed with the direction of GSB's development over the past few months.
Everything is still spinning around the entering to trade.
We are constantly looking for an ideal entry to trade for a few markets.

In GSB we have thousands of settings for everything that some of us do not even understand, but we have no option to set the type of EXIT ... don´t you think it's a little weird?

I don´t know if there is any major edge in your multitimeframe testing method when the exit from the trade is always at CloseD price. I understand that the results will improve a few percent but it is quite logical when EXIT is on the same price on every timeframe...and programmers still work over and over on one and the same ... better entry on just a three markets.

Yes I can ignore your way of multitimeframe testing, but i´m still waiting for implementation of any different exits.
I believe the implementation of only a few other reasonable EXITs than closeD would not be as problematic as multitimeframe testing and would open a new space of strategies on many more markets than only ES, CL, NG.

Anyway, GSB is a great tool, although it is now going through a more complicated way than it might have been.

The pipeline is what I hope to get done in the next week or two.
Exits and additional secondary filters are a high priority, but they are both are a big job that must be done at the same time. So I am trying to get most of the small features that lots of users have request - finished before we start exits. If exits and additional SF filters were a small job, it would be done now. Much time was spend on multi time frame systems, and this is something that has massive significance to nearly (all?) users. The improvement is massive on multi time frame. On crude oil out of sample degradation went from -42 to - 12 %. There are not 1000's of settings, and if whats on the bottom GUI & limited docs are not enough, you can ask. Improved documentation is being working on as well. Regardless Im sorry for your frustration and you will get what your after. If you follow GSB growth in the last year, it has taken massive and consistent leap forward. Many users are asking for lots of things, and I am hoping that most features user ask for get implemented.

zdenekt - 24-8-2018 at 05:59 PM

It's not frustration (I have no problem setting up and using GSB) ... it's just a disillusion that after more than a year of GSB existency and so many different updates we still don´t care about such an important thing as the exits.

I must say again that GSB is a great tool and I really respect your work.
I will try to be patient.

JasonT - 26-8-2018 at 12:33 AM

Quote: Originally posted by zdenekt

... we have no option to set the type of EXIT ...

zdenekt, have you explored the Minute, Stop Loss, Profit Target, Time of Day exits that are available in addition to the close at end of session exists?

zdenekt - 26-8-2018 at 09:50 AM

I'm working mainly on swing strategies so now I can only use stop loss and profit target. The problem is the varying volatility across history.

For intraday trading there are only a few suitable markets and their reactions to different events are very closely correlated.

For swing trading I can use markets from all sectors and I can set up a portfolio as I need.

admin - 26-8-2018 at 04:25 PM

Quote: Originally posted by zdenekt

I'm working mainly on swing strategies so now I can only use stop loss and profit target. The problem is the varying volatility across history.

For intraday trading there are only a few suitable markets and their reactions to different events are very closely correlated.

For swing trading I can use markets from all sectors and I can set up a portfolio as I need.

From memory, GSB as is - can reverse a position, so this is an exit too.
Its not what you want but for now better than just stop and profit target exit

zdenekt - 28-8-2018 at 01:46 PM

Peter,
Based on my research I can not confirm your thesis that reverse position is for now better than stop and profit target.

I tested a lot of markets with just a reverse position exit but I could not find the only strategy that would profitable/tradeable on OOS period. If someone had some succes with it, please show us the results.

While I set up stop and profit target (coexist together with reverse exits) it is possible to find strategy that is profitable on OOS period.

You can see some of my swing strategies on the picture.
All these strategies also passed the Cluster Analysis in Tradestation WFO.

admin - 28-8-2018 at 04:13 PM

Quote: Originally posted by zdenekt

Peter,
Based on my research I can not confirm your thesis that reverse position is for now better than stop and profit target.

I tested a lot of markets with just a reverse position exit but I could not find the only strategy that would profitable/tradeable on OOS period. If someone had some succes with it, please show us the results.

While I set up stop and profit target (coexist together with reverse exits) it is possible to find strategy that is profitable on OOS period.

You can see some of my swing strategies on the picture.
All these strategies also passed the Cluster Analysis in Tradestation WFO.

I'm not saying reverse is better, but its an option as well as stop and PT. The first thing I would check is what secondary filter you have. This is typically one of the most important settings. So try closed-close, and ga with closed-close and ga with close/CloseD
Note the ga S.F. are normalized, while the SF others are not
Can you let me know how this goes. Worked fine on Apple but I didnt do market validation tests on it. Focusing all time on Crude oil right now

zdenekt - 28-8-2018 at 05:15 PM

I have to go sleep so very shortly for now.
Yes, filter is most important element in GSB strategies.

I think that none of my swing strategies use CloseD in any form.
I found that CloseD in swing strategies only works on In-sample period but OOS period always failed. This applies to both the indicator and the filter.
So I often uncheck CloseD and do not using it at all.

For now, I only work with futures markets.
Stocks or ETFs don´t tried yet.

admin - 28-8-2018 at 05:18 PM

Quote: Originally posted by zdenekt

I have to go sleep so very shortly for now.
Yes, filter is most important element in GSB strategies.

I think that none of my swing strategies use CloseD in any form.
I found that CloseD in swing strategies only works on In-sample period but OOS period always failed. This applies to both the indicator and the filter.
So I often uncheck CloseD and do not using it at all.

For now, I only work with futures markets.
Stocks or ETFs don´t tried yet.

Stocks and eft should be fine overall. I have never seen closed filters not work on any market, but I havnt looked at all of your markets before. When you do a support ticket upload, I can look further

Carl - 29-8-2018 at 04:01 AM

Quote: Originally posted by zdenekt

Hi Zedenekt,

The four equity lines on wheat, corn, sugar, cotton and belong to swing strategies that were developed by GSB?

Thanks.

zdenekt - 29-8-2018 at 04:02 AM

Sorry for my english.
I found that swing strategies with CloseD indicator or filter can have positive results on In-sample period but OOS period has always had negative results.

admin - 29-8-2018 at 06:02 AM

Quote: Originally posted by zdenekt

Sorry for my english.
I found that swing strategies with CloseD indicator or filter can have positive results on In-sample period but OOS period has always had negative results.

Under help, do a support ticket and I will look into this

zdenekt - 29-8-2018 at 07:05 AM

I'm talking about robustness of CloseD filter in swing strategies. I did not said it's bug in GSB.

admin - 29-8-2018 at 02:47 PM

Quote: Originally posted by zdenekt

I'm talking about robustness of CloseD filter in swing strategies. I did not said it's bug in GSB.

Agreed, but I want to look into this issue regardless

cotila1 - 2-9-2018 at 01:00 AM

zdenekt, sorry not sure if understood this: your equity lines in pics (corn etc.) are from systems built by GSB?
thanks

Quote: Originally posted by zdenekt

zdenekt - 2-9-2018 at 02:38 AM

Yes, these equity lines are from systems built in GSB.

On the picture you can see one of my potential portfolio of 11 strategies (9 of them are built in GSB).

Portfolio sample - 11 strategies.jpg - 146kB

cotila1 - 2-9-2018 at 12:10 PM

Oh sounds great!
If you dont mind could u mention the 9 symbols you built in GSB?

When does the OoS start from? Is that OoS the so-called test period or is it an unseen period by the software?

thank you so much

Quote: Originally posted by zdenekt

Yes, these equity lines are from systems built in GSB.

On the picture you can see one of my potential portfolio of 11 strategies (9 of them are built in GSB).

admin - 6-9-2018 at 03:46 PM

I ended up with 900 CL systems, which all passed verifcation profit factor 1.8, pearsons 0.95 on the 11 bar intervals 25 to 35 min. A hard job to narrow this down to get fewer systems.
I then added 24 and 26 minute bars, and added 20 and 40 minute bars to verification.
This dropped things down to 300 systems.
I then though, would be good for portfolio analysist pro to sort out the best combination of say 10 systems.
However GSB couldnt do bulk export of all systems to PA.
This feature has now been added.
Sorry for the delay in new releases this week. Interval multiplier has been added. So you can now just use 1 minute bars and multiple by 30 etc
You can also clone bar setups. ie if you have CL,NG,HG,RB 30 minute bar, you can clone and change the interval to 29 minute.
This feature however has at least one bug, and a danger that TS <>GSB in some circumstances. Hence I didnt release it.
If anyone is desperate for the above feature, send me a email & I will supply

admin - 11-9-2018 at 08:58 PM

Most of us have made a great GSB system and put it into TS. We do however loose the GSB format of the system if we are not careful.
Why we might want this is to apply new GSB features to an older system, or to walk forward.
There is some binary info at the end of the TS/MC code of GSB, that can be converted back to a GSB format system
This feature will be in 49.07 release onwards.

admin - 11-9-2018 at 09:03 PM

Another useful feature, is the stats function a&b can be applied to only the systems that are stored on favorites tab.
It means you can test the stats of say systems that have verified on all markets, without having to delete all the non verified markets.
Useful as you might delete some systems with metrics, and turn out to regret that.

admin - 18-9-2018 at 08:34 PM

A new release will be out for trial users by the end of the month. Old code will have expired too.
We now have CLoseDBPV (Big point value) which makes closeD-Close redundant. Much better for when we you compare multiple markets that have very different point values. Lots of small tweaks and features have been added. Much better use of ram when nth / dates tests are used.
There now is a job que on the right side gui for NTH, dates and market verification. (49.09 onwards)
A help file. (not complete) has been added under help. (top menu)

cotila1 - 18-9-2018 at 11:48 PM

I like this new CloseD (CLoseDBPV) . That means that if I use the close-closeD(1) as SF for ES and I want to verify on nq, then i should not have big difference/issue by using this new CLoseDBPV?
So the old ''CloseLessPrevCLoseD'' will have new name in SF dropdown menu?

Quote: Originally posted by admin

A new release will be out for trial users by the end of the month. Old code will have expired too.
We now have CLoseDBPV (Big point value) which makes closeD-Close redundant. Much better for when we you compare multiple markets that have very different point values. Lots of small tweaks and features have been added. Much better use of ram when nth / dates tests are used.
There now is a job que on the right side gui for NTH, dates and market verification. (49.09 onwards)
A help file. (not complete) has been added under help. (top menu)

admin - 18-9-2018 at 11:51 PM

Quote: Originally posted by cotila1

I like this new CloseD (CLoseDBPV) . That means that if I use the close-closeD(1) as SF for ES and I want to verify on nq, then i should not have big difference/issue by using this new CLoseDBPV?
So the old ''CloseLessPrevCLoseD'' will have new name in SF dropdown menu?

Quote: Originally posted by admin

correct on all points but dropdown box will have closed and closedBPV

cotila1 - 18-9-2018 at 11:54 PM

Quote: Originally posted by admin

Quote: Originally posted by cotila1

Quote: Originally posted by admin

correct on all points but dropdown box will have closed and closedBPV

So If i want to use close-closeD(1) as SF for ES building and I want to verify on nq and avoid the BPV issue between es and nq, should I choose closed or closedBPV for building?

admin - 19-9-2018 at 01:28 AM

Quote: Originally posted by cotila1

Quote: Originally posted by admin

Quote: Originally posted by cotila1

Quote: Originally posted by admin

correct on all points but dropdown box will have closed and closedBPV

So If i want to use close-closeD(1) as SF for ES building and I want to verify on nq and avoid the BPV issue between es and nq, should I choose closed or closedBPV for building?

Use bpv, there is no need to use close-CloseD any more. Beta 49.09 uploaded in private forum shortly

admin - 21-9-2018 at 06:02 PM

49.12 has comments in data table. They are only for the user, not gsb.
There always is lots of confusion over session times.
Your TS should run local time. This is because if you have a data 2 with different time zone to data1, TS&MC wont allow it.
The session time on your chart is exchange time, but the exported data is in your local time.
Moc time should = local time last bar of day, not exchange time.
Note also cash indices have changed point value to what is on the futures contract. This is needed for closeDBPV (big point value) to work best

admin - 1-10-2018 at 01:11 PM

In the beta section (GSB paid members) 49.20 was released today. Lots of long awaited features.
One of the best things in GSB is the ability to build 1000's of systems. Then tweak things to see what improves the out of sample results. We were using every second day as un seen to get a out of sample picture. Then we look at how much out of sample degrades compared to in sample. Well now we can choose say 3 days to trade, 1 day to not trade. This means greater in sample period - so we get more trades to form a more valid opinion. We can then get just as many trades out of sample by twice the amount of trades. (Before we traded 1 day, then 1 day out of sample) So now the statistics are normalized to take into account that you have less trading days.
We can also have unlimited custom dates for training, test and validation.
There is now a job que to submit changes in the dates or nth so changes in dates and nth can be done in one pass, not two.

kelsotrader - 2-10-2018 at 02:25 PM

Quote: Originally posted by admin

49.12 has comments in data table. They are only for the user, not gsb.
There always is lots of confusion over session times.
Your TS should run local time. This is because if you have a data 2 with different time zone to data1, TS&MC wont allow it.
The session time on your chart is exchange time, but the exported data is in your local time.
Moc time should = local time last bar of day, not exchange time.
Note also cash indices have changed point value to what is on the futures contract. This is needed for closeDBPV (big point value) to work best

Excellent Explanation: I always add the exchange session time onto the file name when exported so that the GSB code will record the time in its code so as to keep a permanent record on the strategy.

admin - 18-10-2018 at 06:45 PM

Soon we will have market degradation score per system.
So lets say you build 5000 CL systems on 29,30,31 min bars.
You can then verify on 20,24...36,40 min bars. (12 different bar intervals)
You would expect the 20,24...36,40 min bars to give a bit worse performance than the 29,30,31.
Then those systems that get verification score of say 10 up, or even 12.
You can then choose the system where the verification degradation is the highest.
Shown here the verified bars had higher performance than the 29,30,31 that the system was build on.
The out of sample (alternate nth day) also had higher results than in sample.
I would say this would be very unusual.
800 systems were tested to get this result.

admin - 19-10-2018 at 04:47 PM

We can now turn off and on specific groups of curves, and the legend. This stops the graphs being too cluttered.
In build 49.44 onwards

cyrus68 - 20-10-2018 at 01:37 AM

I agree entirely. I think the information on the graph can be revised to reduce clutter and be more relevant. Instead of all the line items showing the metrics for each verification dataset (whether it can be turned on or off), we could have a single one showing the average metrics for all the verification datasets. This will allow us to make a direct comparison with the average result of the datasets that were used in the build. Is this what the VSS score is showing?

Having reduced the clutter, it is important to include line items for the metrics of EACH of the datasets that were used in the build – in this case, 29 30 31 min. I would like to see both the IS and OOS results. This is essential information when we want to run WF. I don’t automatically run WF on the central bar size and, even if I do, I want to see its metrics – not just the average IS and OOS for all of them.

admin - 21-10-2018 at 03:12 PM

Quote: Originally posted by cyrus68

I agree entirely. I think the information on the graph can be revised to reduce clutter and be more relevant. Instead of all the line items showing the metrics for each verification dataset (whether it can be turned on or off), we could have a single one showing the average metrics for all the verification datasets. This will allow us to make a direct comparison with the average result of the datasets that were used in the build. Is this what the VSS score is showing?

Having reduced the clutter, it is important to include line items for the metrics of EACH of the datasets that were used in the build – in this case, 29 30 31 min. I would like to see both the IS and OOS results. This is essential information when we want to run WF. I don’t automatically run WF on the central bar size and, even if I do, I want to see its metrics – not just the average IS and OOS for all of them.

VSS is showing the metric for the individual time frame . market -not the average. I will look into "see both the IS and OOS results". This will be even more info in the GUI.

admin - 31-10-2018 at 10:32 PM

Additional work on exits has started about 2 weeks ago. Hence the lack of other updates the last few weeks.
I'm nearly finished the next video on CL, with significant refinement of methodology.
The next release will appear in the beta forum for paid users.

admin - 26-11-2018 at 05:39 PM

In the beta forum, we now have trailing stops, atr stops, exits after x bars (we has exits after x minutes some time ago)
Additional exits are being debugged now.
The new video is out. https://www.youtube.com/watch?v=HDeJpONE090
verification of systems done in the workers, instead of only in GSB managers.
Soon we should have counter trend oscillators which should result in counter trend systems

admin - 30-11-2018 at 01:00 AM

GSB counter trend oscillator released in the private forum today.

Very Impressed

JasonT - 10-12-2018 at 01:10 AM

Hi Peter,

I just wanted to say how impressed I am with the 50.25 beta version of GSB. I'm finding it much easier to use than the 49.xx series, especially the way you've incorporated the IS/OOS degradation statistics - fantastic.

I've gone through the last 3 you tube videos a few times over the weekend and have got a much better idea of how to perform degradation testing. It took me a while to piece it all together. Now that I understand what you are doing and why, it is straight forward. Good work.

Videos here:
a) Genetic System Builder Market testing / validation 22 May 2018
b) GSB Market Validation video2 2 Jul 2018
c) Genetic system builder, new features to get better out of sample results 1 Nov 2018

JasonT - 10-12-2018 at 01:20 AM

Quote: Originally posted by JasonT

I just wanted to say how impressed I am with the 50.25 beta version of GSB.

I'm also a fan of the User Tasks section in the resource monitor panel on the right. It's great to see progress against a change of Nth day mode or Verification calculation.

admin - 31-12-2018 at 12:19 AM

Short term updates.
Coming soon.
Subtle improvements to the allocation of GSB cloud.
Export of all the metrics of say just 30 minute bars, even if we build on 29,30,31
various bug fixes

Not to soon we start on new/additional secondary filters.
GSBramv2 is a bit over a week away. This will allocate your idle cpu to the general cloud resulting in more cloud for all.

admin - 12-2-2019 at 10:15 PM

Resource manager (RM) is running well, and there now is decent cloud power for all. RM is contributing unused CPU power from GSB users, and I have 4 high end machines that often contribute to the cloud.

Updates to RM are planned. Ie run on computer startup. Start x workers on startup, rather than starting just 1.

Additional price, time filters are being built right now. We will start with a small library just to test that it all works, then increase the features.
This hopefully will make a much more powerful GSB. BEST case the first beta build will be ready at the end of the week.

admin - 17-2-2019 at 11:33 PM

This is what the next build of GSB looks like. Likely these new groups in cyan will be combined into one function each.
This is not a exhaustive list, just a few added for proof of concept.

admin - 17-4-2019 at 08:14 PM

GSB 2.0
I have been asked about GSB 2.0

There are a number of things that need to be revamped, and a lot of that is to make user errors less probable.
Richard Bransons space shuttle crashed, just because someone pressed a button at the wrong time. GSB issues are much more minor, but this section needs a revamp.

The first version of this is basically going to have the secondary filters reduced to closeD variants,
and the normalized and non normalized are all going to be in the same set of filters.
Currently you have non normalized CLoseD.. or GA (which are all normalized) with closeD and 37 or so indicators.
(Normalization is where we adjust the last 100 bars of data to be in a range of -100 to +100)

The number one issue users make is wrong secondary filter (SF). It is critical to get this right.
Some users also add all 38 SF. This results in much slower build of systems, and the systems that dont chose a Closed SF, work much poorer.
So SF are going to be soft coded to the contract used. So if a user does nothing, the setting will be correct.
So this section is simplified, and secondary filters could ALL be chosen genetically. Currently this cant be done.

We will keep compatibility with old systems in new managers of GSB 2.0, but we will loose optimizer setting compatibility.
app settings will be the same.

Opt settings needs the return of secondary filter closed parameters, but a different set of values if closed-close vs closed/close.
These sorts of things bit too many users, which is why we hid them and made them auto detect. Other uses wanted control of this variable.
Likely we also need to have the indicator weights linked to what operands we use.
ie if operand * is used, we use weight -1 to 1 step 2 and the others set to 1
if addition is used, we might use -100 to 100 step 10 etc

For new users, the over simplified GSB system format is

result=indicator1*weight1{operand = */+}indicator2*weight2{operand = */+}indicator3*weight3{operand = */+}
secondary filter =closeD(1)-close
if result >0 and secondaryFilter=true then buy..

Im also hoping to add the ability for gsb to make the binary files that EWFO and ts wfo.exe makes.
My wild guess is GSB is going to be 10 to 100 times faster to make these.
EWFO has a number of features that gsb wf doesnt have. Cluster analysts, semi automatic fitness detection.
Wf efficiency (where out of sample has to be reasonable relative to in sample)
These are not essential features, but a number of users have expressed interest in this.

The big area of GSB growth will be in tertiary filters. Tertiary means 3, so the GSB over simplied system format will have added

if result >0 and secondaryFilter=true and Tertiaryfilter=true
then buy..

There is massive scope in Tertiary filters, and we will do a few things, test for bugs, then continue to expand

coccigelus - 18-4-2019 at 03:03 AM

Very interesting the new many directions GSB is taking. I think the most interesting feature will be the implementing of the price pattern and could really open a lot of new doors and non correlated systems including to avoid some critics things involved with oscillators. That area should be easily expanded later increasing the number of candles analyzed.

Regarding the tertiary filters I am skeptical for a couples of reasons.

Kudos to the work You are doing on the opt/fitness/normalization criteria. That is really important and really needed including increase flexibility for the users.

THX for feed back.

admin - 20-5-2019 at 05:14 AM

Whats almost finished is the ability to walk forward gsb systems in GSB, but send the output to EWFO
https://trademaid.info/hugo/ewfo.html

GSB WF is fast, but EWFO WF is much more powerful. This gives you the best of both worlds.

cotila1 - 20-5-2019 at 08:29 AM

Quote: Originally posted by admin

Whats almost finished is the ability to walk forward gsb systems in GSB, but send the output to EWFO
https://trademaid.info/hugo/ewfo.html

GSB WF is fast, but EWFO WF is much more powerful. This gives you the best of both worlds.

that's a very clever feature. thank you

admin - 20-5-2019 at 11:27 PM

This is whats new in build 52.32
Macros instant cancel implemented.
Macros logging (minimal to text file) implemented.
The sorting and top added to favorites macro.
Sound macro implemented.
Email macro implemented.
FavoritesReset macro and right-click button implemented.
StatsReset macro and right-click button implemented.
Stats G-H implemented.
WF Params. (separate from Param.) implemented.
Subtract operator added.
Find (system by ID) button implements.
Improved GUI responsiveness, CPU and RAM on managers (by disabling backtest on new systems download).
Allowed Nth invert (and similar tasks) to run multi-threaded while optimization is active.
Multiple fixes in cloud WFs.
Selected Rows/Cols/Cells implemented.
Minor GUI updates and fixes.

avatartrader - 21-5-2019 at 08:01 PM

Hi Peter,

I have not had a chance to update and test just yet, but by "WF Params. (separate from Param.) implemented", are you referring to the ability to save and restore WF params when saving systems that have been walked forward, or something else?

admin - 21-5-2019 at 08:06 PM

Quote: Originally posted by avatartrader

Hi Peter,

I have not had a chance to update and test just yet, but by "WF Params. (separate from Param.) implemented", are you referring to the ability to save and restore WF params when saving systems that have been walked forward, or something else?

This is wf with different weights to the weights used to build.
Interesting I did a test of 50,000 systems weights 0.5,2 step 0.25
but did wf with wigths -1 to 2, step 0.25
operator + only
Interesing that no weights of 0 or below were chosen.
This is a health sign. If weight of 0 of below was chosen, then the indicator it refers too would be considered bad or redundant logic.

Saving of WF system paramaters is in the short to meduim term pipeline, but not sure when.

GSB wf to ewfo is done and likely released within two builds

admin - 5-6-2019 at 10:15 PM

Walk forward from GSB to EWFO has been done.
Much faster back end processing of WF jobs implemented.

IN version 52.72, a lot more work done in macros.
The comments field, else file name will appear in the macro mouseover.

Resource monitor will soon be Auto config to maximize whatever ram the user has.

admin - 12-6-2019 at 02:06 AM

We now have a graph that shows the ration of GSB calculations per system.
Low is better, and it will increase over time if your getting a lot of duplicates.

We also have macro log. Green is where the macro is right now.

admin - 12-6-2019 at 02:06 AM

Coming later is walk forward with artificial intelligence.

cotila1 - 12-6-2019 at 02:12 AM

cool feature!
more and more progress in gsb, very good.

I was wandering what are the main remaining features to be implemented yet?
I think one of the main one was SF? which I think is now completed?

when the beta version is planned to be over?

admin - 12-6-2019 at 08:32 PM

Quote: Originally posted by cotila1

cool feature!
more and more progress in gsb, very good.

I was wandering what are the main remaining features to be implemented yet?
I think one of the main one was SF? which I think is now completed?

when the beta version is planned to be over?

There is years of work to go in GSB.
There are now two programmers on GSB, so a job que for each.
Primary programmers task.
Still playing with GSB architecture. IE we added power , log functions + other stuff we experiment with. Some of these little tweaks give significant out of sample improvements when you add them together.
New indicators to be added.
.... So all the new features we suddenly decide we want push the secondary filters further out. IM hoping that will start soon. It will lead to GSB 2.0 which is going to have some greater intelligence to avoid user error, but also have the option to tweak things. The secondary filters (likely we call them teritary filters) is a massive task to be rolled out in stages.

Secondary programmer.
More in macros. ie add to favorites certain vs and vss score, macro loging.
C/US graphing. This is going to have more work done. Its given me significant incites into how GSB is crunching things, and ways to improve.

Then ratio of degradation in sample to out of sample per system, not just the systems as a collective group.

AI walk forward. Thats a huge task.

SO I guess we get stable releases for the trial users, and constant beta build for the GSB purchases with the latest tweaks and features.
Trial users only get updates every 3 months, beta users sometimes a few per week. It varies a lot.

Hope that helps.

cotila1 - 13-6-2019 at 01:16 AM

Yes, it helps indeed and thanks for showing this!

I see an exciting and challenging roadmap
Im sure to see great results.

Quote: Originally posted by admin

Quote: Originally posted by cotila1

admin - 25-6-2019 at 12:33 AM

AI walk forward work has started.
We have a new entry type,
improved filtering of duplicate systems.
53.40 build due this week, GSB purchasers.
Includes bug fixes, and numerous other small tweaks