GSB Forums - Powered by XMB 1.9.11

Update to GSB methodology. A must read, the backpacker and the Art of war by Sun Tzu

Pages: 1 .. 6 7 8 9 10 .. 12

admin - 28-3-2021 at 05:33 PM

@RWS
As carl said, familes is doing this in a more crude way.
There are a number of issues.
Likely a WF should be done first, as the original system parameters might be way off the wf final parameters
Your example shows 1 parameter.
Lets say there are 10 parameters, and we shift all of them.
Do we shift one at time, or all 10 at once?
Using the verify feature on different bar intervals indirectly does this.
This is because there are 13 bars in 830 to 1500 day 30 min
but if its 29 min, then its 14 bars.
31 min 12 bars. So you are indirectly changing the parameter length.
GSB also has verification where you can compare other markets to what you built on.

"One way to predict the SP500 index is to count the number of buy/sell signals of it's stocks which seems to work reliable."
If your willing to put the time into loading 500 symbols in GSB, this could also be done as is with no changes needed to GSB

rws - 28-3-2021 at 06:39 PM

Hi Peter,

There are usually much more parameters and Stratasearch shifts them all together 1 up for 1-10 and 0.1 for 0.1-1. I think there is room for improvement. It could be better to have all parameters separately tested but that would take much CPU time so this is a bit crude way also. Also, it could be better if all parameters are not shifted 1 step but a certain percentage and then rounded if the indicator needs integers.
Still I find improvement OOS having the parameters shift on while building. Building on 450 tickers and having a low drawdown the last 20 years and having good and broad parameter stability I did not see the need of walkforwarding anymore. It takes a long time to build but the result is often very similar OOS as IS.

You can compare till some extend to bar intervals. For example, changing from 13 to 14 minutes means all parameters should change times 13/14.

But changing some parameters from 5 to 6 and 7 and some
from 11 to 12 and 13 has a different scaling. Also, in the example below the number 150 being the highest atr of 150 bars is not changed while all other are.
Also the way you directly optimize on a custom criteria that has rules in it to have broader peaks for custom metrics is not directly possible with only changing bar intervals and/or picking the top of a filtered list of results. In the way it is done in stratasearch it is a more direct way in guiding the genetic optimizer to search for systems that are less dependent on changes in parameters which often translate in being less sensitive for market changes and better OOS performance. A peak optimization should always be avoided.

Entry
// MACD: 1
(crossabove(macd(), mov(macd(), 5, exponential))) and

// Average True Range - Low Stock Periods
(atr(30) < high(atr(30), 150) * 0.8) and

// Dynamic Momentum Index - Sector Periods: 1
(sector("S&P 500 Index", dmi()) < 76)

Exit
// DEMA: 1
(crossbelow(dema(close, 19), dema(close, 21))) or

// Average True Range - Low Stock Periods
(atr(30) > high(atr(30), 150) * 0.8) or

// Dynamic Momentum Index - Sector Periods: 1
(sector("S&P 500 Index", dmi()) > 76) or

// Gap Down Sell
(IsGapDown())

Quote: Originally posted by admin

@RWS
As carl said, familes is doing this in a more crude way.
There are a number of issues.
Likely a WF should be done first, as the original system parameters might be way off the wf final parameters
Your example shows 1 parameter.
Lets say there are 10 parameters, and we shift all of them.
Do we shift one at time, or all 10 at once?
Using the verify feature on different bar intervals indirectly does this.
This is because there are 13 bars in 830 to 1500 day 30 min
but if its 29 min, then its 14 bars.
31 min 12 bars. So you are indirectly changing the parameter length.
GSB also has verification where you can compare other markets to what you built on.

"One way to predict the SP500 index is to count the number of buy/sell signals of it's stocks which seems to work reliable."
If your willing to put the time into loading 500 symbols in GSB, this could also be done as is with no changes needed to GSB

admin - 28-3-2021 at 08:02 PM

@rws. Need to think on this. The ideal is all parameters are shifted individually and results summed up. Lot of work to do this, but likely you could do this in ts with one system, exporting results to .txt file
and sume them all in excel.
You can also add 500 es stock, or simpler 100 NQ stock to test your proposal. It makes some sense.
Again there are problems with practicalities. Stock get rotated in and out of the indexes. IE Tesla

Carl - 29-3-2021 at 01:34 AM

I am just thinking out loud here... could something like this be beneficial?

What if we would take GSB variable "result" and calculate its correlation to "result" with parameters 30% larger and 30% smaller?

So for example, a GSB strategy was built with 3 primary indicators (PF).
Length PF1 10
Length PF2 100
Length PF3 40
Result = X

parameters increased by 30%
Length PF1 13 (= 10+30%)
Length PF2 130 (= 100+30%)
Length PF3 52 (= 40+30%)
Result = Y

parameters decreased by 30%
Length PF1 7 (= 10-30%)
Length PF2 70 (= 100-30%)
Length PF3 28 (= 40-30%)
Result = Z

Correlation between
X and Y
X and Z

If correlation is high, then the variable result is stable across the original GSB values and +30% and -30%. So the strategy is robust?

admin - 29-3-2021 at 03:26 AM

@carl, Its still a stab in the dark compared to multiple tests with all parameters going up & down. It might be better than nothing but there is room to improve this.
Forcing say 100+ members of one family would be better and do-able on GSB with no updates.
Build a system, then reduce the indicator set to force the 3 indicators on, and build 100+ systems
Then look at the oos of all 100 systems

rws - 29-3-2021 at 06:35 AM

It is hard to select a system from a list which has it's parameters more
or less optimized in the middle of a broad peak that is configurable.

Quote: Originally posted by admin

@carl, Its still a stab in the dark compared to multiple tests with all parameters going up & down. It might be better than nothing but there is room to improve this.
Forcing say 100+ members of one family would be better and do-able on GSB with no updates.
Build a system, then reduce the indicator set to force the 3 indicators on, and build 100+ systems
Then look at the oos of all 100 systems

rws - 29-3-2021 at 02:05 PM

In stratasearch in this case I have a filter which rejects

Y/X<0.9 and Y/X>1.1 or a bit bigger
The second condition is necessary because if Y/X is also
part of a customcriteria formula it could benefit the systems
that have a big peak next to it's current values which is also
suboptimal but there is also something positive to that.
I wonder what happens if you choose a parameter setting
in it's through instead of a peak. So left and right of the
parameter setting it has higher performance.
If that parameter setting would have acceptable performance
it could maybe also have some positive effect OOS.

Can what you suggest be automated?

Quote: Originally posted by Carl

I am just thinking out loud here... could something like this be beneficial?

What if we would take GSB variable "result" and calculate its correlation to "result" with parameters 30% larger and 30% smaller?

So for example, a GSB strategy was built with 3 primary indicators (PF).
Length PF1 10
Length PF2 100
Length PF3 40
Result = X

parameters increased by 30%
Length PF1 13 (= 10+30%)
Length PF2 130 (= 100+30%)
Length PF3 52 (= 40+30%)
Result = Y

parameters decreased by 30%
Length PF1 7 (= 10-30%)
Length PF2 70 (= 100-30%)
Length PF3 28 (= 40-30%)
Result = Z

Correlation between
X and Y
X and Z

If correlation is high, then the variable result is stable across the original GSB values and +30% and -30%. So the strategy is robust?

admin - 30-3-2021 at 12:48 AM

@rws
im open to the idea, but need to think on how to do it. This might be a lot of work.
As I said previously we have a little similar concept using families, but the ranges are more chosen by GA, then a fixed % up and down
Here you can see 6 systems with 6 or 5 members in each system. All are profitable with reasonable metrics. I like family 4 in light blue best

rws - 30-3-2021 at 10:32 AM

I see it causes a huge number of calculations if all parameters combinations should be checked. Suppose there are 3 steps up and 3 steps down and in total 10 parameters that all need to be tested.
In that case there are 6^10 combinations to recalculate which is far too much. So the simple way of adding all indicators 1 step or rather a % seems to me the only practical way which is not perfect but it could help.
In this case there are only 6 steps.

I don't know what optimizer you use but I understand there are optimizers that are able to avoid peaks, maybe that brings more.
Amibroker says they use these and the owner says it searches for wide areas.
https://www.amibroker.com/guide/h_optimization.html
I still find it is a peak optimizer. The problem is there are no settings how broad a peak should be and if the system only has a peak in a narrow area you can't exclude it. But these type of optimizers have ways to efficiently search for peaks and it could be that changing that code could bring more finding broad areas or excluding narrow peaks than trying many combination to confirm afterwards.

You wrote
As I said previously we have a little similar concept using families, but the ranges are more chosen by a, then a fixed % up and down

What do you mean with "the ranges are more chosen by a,"

I will soon start with GSB again. Have been spending a lot of time
in Stratasearch.

Quote: Originally posted by admin

@rws
im open to the idea, but need to think on how to do it. This might be a lot of work.
As I said previously we have a little similar concept using families, but the ranges are more chosen by a, then a fixed % up and down
Here you can see 6 systems with 6 or 5 members in each system. All are profitable with reasonable metrics. I like family 4 in light blue best

rws - 30-3-2021 at 10:37 AM

I have been thinking, if you discard the dependecy of parameters
and still do a parameter check for every parameter to avoid peaks
then you can reduce the number of calculation to 10x6= 60 which
is doable. I think it would be much better then the other alternatives.
If all used parameters are not in a peak area it is hard to imagine that all together they do form a peak, rather the opposite.

Quote: Originally posted by rws

I see it causes a huge number of calculations if all parameters combinations should be checked. Suppose there are 3 steps up and 3 steps down and in total 10 parameters that all need to be tested.
In that case there are 6^10 combinations to recalculate which is far too much. So the simple way of adding all indicators 1 step or rather a % seems to me the only practical way which is not perfect but it could help.
In this case there are only 6 steps.

I don't know what optimizer you use but I understand there are optimizers that are able to avoid peaks, maybe that brings more.
Amibroker says they use these and the owner says it searches for wide areas.
https://www.amibroker.com/guide/h_optimization.html
I still find it is a peak optimizer. The problem is there are no settings how broad a peak should be and if the system only has a peak in a narrow area you can't exclude it. But these type of optimizers have ways to efficiently search for peaks and it could be that changing that code could bring more finding broad areas or excluding narrow peaks than trying many combination to confirm afterwards.

You wrote
As I said previously we have a little similar concept using families, but the ranges are more chosen by a, then a fixed % up and down

What do you mean with "the ranges are more chosen by a,"

I will soon start with GSB again. Have been spending a lot of time
in Stratasearch.

Quote: Originally posted by admin

admin - 30-3-2021 at 03:33 PM

@rws. I fixed the typo. Gramer checker changes ga to A.
So what your saying is each parameter in increased / decreased 1 at a time?
Another option is you do x tests of randomized parameter changes.

Were you aware how simple it is to add random noise to GSB data too?

bartek - 30-3-2021 at 04:12 PM

First time I run GSB for NQ and got some nice results.

admin - 30-3-2021 at 04:22 PM

Quote: Originally posted by bartek

First time I run GSB for NQ and got some nice results.

Thats great.
A few comments.
poor results pre 2007 in really common. I start at 20070701
What entry / secondary filter type did you use?
The next release of GSB has a terteiry filter of not normalized CloseLessPrevCloseD
which works well with closslesshighlow3 with nq

bartek - 30-3-2021 at 04:26 PM

Quote:

Entry mode: cross, SF: CloseToHighLow3v2,

admin - 30-3-2021 at 05:46 PM

@barteck
adding a closedFilter likely will help your NQ system
Build 61.82 onwards has this as tertiary filter. Its not released yet

inputs:offset(0);//optimize this

If time < 1500 then
Begin
// Buy/Sell
If (decision = 1 Or decision = 2) And sfDecision = 1 Then
Begin
IF CURRENTCONTRACTS>0 OR CLOSE+OFFSET>CLOSED(1) THEN

Buy ("Long Entry4") 1 contracts this bar on close;
flag = 1;
End
Else If (decision = -1 Or decision = 2) And sfDecision = -1 Then
Begin
IF CURRENTCONTRACTS>0 OR CLOSE-OFFSET
SellShort("Short Entry4") 1 contracts this bar on close;
flag = -1;
End;
End;

bartek - 31-3-2021 at 01:43 AM

@Peter

With tertiary filter do you still use 3 or only 2 indicators?

rws - 31-3-2021 at 02:28 PM

@peter
Yes one at a time and being able to acces the result of the + shift and - shift seperately.

I think with a parameter space in this example of 6^10 you will get not so much idea of where flat areas are based on a list of results. Especially since they are all optimized result you would probably find results of other peaks in the parameter space.
You would have some indea of maximum and minimum values but I can't see how you could pick the systems that are stable arounds it initial startpoint.

But if you set a limit on how much higher or lower the profit should as a result of a upshifts or downshifts of parameters then you can filter systems that are in a peak area. By having the resultnumbers like profit, profit/drawdown etc for the up and downshifts you can define the shape of the curve as a function of the parameter changes.

Quote: Originally posted by admin

@rws. I fixed the typo. Gramer checker changes ga to A.
So what your saying is each parameter in increased / decreased 1 at a time?
Another option is you do x tests of randomized parameter changes.

Were you aware how simple it is to add random noise to GSB data too?

admin - 31-3-2021 at 03:52 PM

@rws
i will pass onto the programer
if you want "Yes one at a time and being able to acces the result of the + shift and - shift separately. "
Do you mean the + results, and the - results or all + & all - results

this means the GUI is going to be messy, or some sort of separate tab is needed.
you cant do this to a 100 systems at once with such a clean gui

bartek - 1-4-2021 at 12:58 AM

And another first run. RTY this time.

admin - 1-4-2021 at 01:10 AM

@bartek, thats impressive though the ave trade is low unless you have added slippage and commission.
Im suspecting closelsshighlow3 works on many markets

bartek - 1-4-2021 at 01:44 AM

Quote: Originally posted by admin

@bartek, thats impressive though the ave trade is low unless you have added slippage and commission.
Im suspecting closelsshighlow3 works on many markets

Results are with commision but without slippage. What is the minimum ave to be set for 300 favs? For gold was set 90.

admin - 1-4-2021 at 01:48 AM

Quote: Originally posted by bartek

Quote: Originally posted by admin

@bartek, thats impressive though the ave trade is low unless you have added slippage and commission.
Im suspecting closelsshighlow3 works on many markets

Results are with commision but without slippage. What is the minimum ave to be set for 300 favs? For gold was set 90.

I'm using the same gold macro for numerous instruments. should be ok as is.

Carl - 1-4-2021 at 06:57 AM

Quote: Originally posted by rws

I have been thinking, if you discard the dependecy of parameters
and still do a parameter check for every parameter to avoid peaks
then you can reduce the number of calculation to 10x6= 60 which
is doable. I think it would be much better then the other alternatives.
If all used parameters are not in a peak area it is hard to imagine that all together they do form a peak, rather the opposite.

Hi rws, I agree.

If we are looking for a stable total outcome, this means every component in itself must be stable as well. Otherwise it could have been a coincidence.

Maybe it is enough to test every single indicator with the same indicator but changing the parameter values by -30% and +30%?

So suppose GSB comes up with a strategy that includes RSI(X).
Just compare the two couples RSI(X) RSI(X-30%) and RSI(X) RSI(X+30%).

Analysis results for example correlation and/or T-test?

For example, look at the correlation between:
RSI(8) and RSI(11) versus ROC(8) and ROC(11)

RSI stability looks good. Correlation 98.6%.
ROC stability not as good. Correlation only 85.6%

admin - 2-4-2021 at 06:06 PM

Update for the week.
Its been very successful in development.
I have modified any indicator cross entry type (AIC). Called AIC2
Im not getting better system building over AIC, but I have found a way to use it regardless.
Later I'm going to explain a lot about AIC entry type. In the past I have been a bit negitve towards it as it has some weakness (that is easy to over come)
There are things nearly everyone doesn't know about AIC that need to be explained.
What I have done post system building is to allow all entries to trade.
This means you can get up to 3 trades in the same day, though 3 is very rare.
I enclose a system report. Included is 1 tick slippage and $2.40 per side slippage.
Give me an hour and I will post more.
This is combine with filters. (Out in the next build of GSB)
The filter I'm using is closelessCloseDbpv. Its VIP to understand this is NOT normalized, while everything else in GSB is normalized. Normaliztion is an indicator is adjusted to be in the range of -100 to 100 regardless of its value

Attachment: Login to view the details

admin - 2-4-2021 at 06:47 PM

here are results of the es settings on other markets. zero slippage used here. all data pre 2007-7 is also out of sample.

bartek - 3-4-2021 at 03:12 AM

Quote: Originally posted by admin

Update for the week.
Its been very successful in development.
I have modified any indicator cross entry type (AIC). Called AIC2
Im not getting better system building over AIC, but I have found a way to use it regardless.
Later I'm going to explain a lot about AIC entry type. In the past I have been a bit negitve towards it as it has some weakness (that is easy to over come)
There are things nearly everyone doesn't know about AIC that need to be explained.
What I have done post system building is to allow all entries to trade.
This means you can get up to 3 trades in the same day, though 3 is very rare.
I enclose a system report. Included is 1 tick slippage and $2.40 per side slippage.
Give me an hour and I will post more.
This is combine with filters. (Out in the next build of GSB)
The filter I'm using is closelessCloseDbpv. Its VIP to understand this is NOT normalized, while everything else in GSB is normalized. Normaliztion is an indicator is adjusted to be in the range of -100 to 100 regardless of its value

Peter, results are really impressive! Awaiting your explanations

rws - 5-4-2021 at 04:48 PM

I rather look at it just like it is an XY graph where the
parameter is defined on the X and the Y is the profit, or
drawdown or some other common metrics.
By defining the max and min profit for the shifts, you can
define how the profit curve has to depend on the parameter change.
You just define how that graph would look like either convex or hollow,
both can have a use.

I think the way it is in Stratasearch is oke but only 1 shift
should not be adding 0.1, 1 or 10 to a parameter but a
predifined% rounded to an integer of that parameter.
And Stratasearch chooses an average of the + and - shift
which is not good. So suppose RSI 10 has 100 k profit and
RSI 9 has 90 k profit and RSI 8 has 110k profit. In this case
the 1 shift result would show 100 k which seems like it has
a good stable area but that is far from the truth.

This is how I define it in SS
1S is the return for one shifted

$V_1SAvgAnnReturn/ $V_AvgAnnReturn > 0.9 and < 1.1
$V_2SAvgAnnReturn/ $V_AvgAnnReturn > 0.8 and < 1.2
$V_3SAvgAnnReturn/ $V_AvgAnnReturn > 0.7 and < 1.3

That helps finding good OS

But since this is an average it should account for both
the -shift and + shift.
This is not possible in stratasearch but just as an example

$V_+1SAvgAnnReturn/ $V_AvgAnnReturn > 0.9 and < 1.1
$V_-1SAvgAnnReturn/ $V_AvgAnnReturn > 0.9 and < 1.1
$V_+2SAvgAnnReturn/ $V_AvgAnnReturn > 0.8 and < 1.2
$V_-2SAvgAnnReturn/ $V_AvgAnnReturn > 0.8 and < 1.2
$V_+3SAvgAnnReturn/ $V_AvgAnnReturn > 0.7 and < 1.3
$V_-3SAvgAnnReturn/ $V_AvgAnnReturn > 0.7 and < 1.3

If a shifted parameter result is not in the range the system is
simply filtered out. In SS you can have up to 5 shifts.

You can either use these statements for a filter and/or
you can use them in a custom formula so it optimizes
on a certain outcome. I optimize them also in the customcriteria
in a way that it searches for a high shifted result.

So for example NPxATx
x$V_+1SAvgAnnReturn/ $V_AvgAnnReturn
x$V_-1SAvgAnnReturn/ $V_AvgAnnReturn
In this case I restrict the maximum value of the shifted
outcomes because otherwise it could find a
low current profit where the shifted profit has a very big value.
You don't want that result either.
There are some more tricks to work around that like having
the shiftedstatement on the power of a smaller than 1 exponent.

Quote: Originally posted by Carl

Quote: Originally posted by rws

admin - 5-4-2021 at 09:19 PM

Im still very open to this idea and its been passed onto the programer to comment, but here is an observation.
ts with wfo use to be able to optimize +=5% and +- 10% on the parameter range or both
of course this means opt takes 3 or 6 times longer.
It was many years since I did this, but I observed it degraded the OOS results.
I tried on quite a few systems.

rws - 6-4-2021 at 02:33 PM

The thing is if you have peaked area with a higher IS profit then you have a higher OS if you manage to get in that peaked area OS.

If market conditions change it could be that a less peaked area performed better OS. But in the less peaked area the OS/IS could be better but IS could have been already less to start with.

If you run several systems and often manage to find at least some in the peaked area OS then I also wonder how much sense it makes.

I also find systems that GSB builds already have a predefined structure where it often already performs very well OS. But still there is often quite some difference between systems IS and OS.

I find when optimizing on 450 stocks, OS is very similar to IS when there is a less peaked area in SS. But SS allows many type of different systems which have complete different characteristics unlike GSB. The priority to have a broad working parameterset might be more important in SS.
GSB works good OS out of the box but compared to SS I really miss the time functions and RAR. Like how many bars the average trade takes and how many bars losing trades take. Of course you can select the systems based on the results but letting it optimize on shorter trades could have found shorter trades initially with the same profit. I suspect that optimizes less efficient for systems that go overnight because I cannot let the GSB optimizer know that a system that takes half the number of bars to have the same profit is better. The custom criteria does not allow me to let that know to the optimizer.
Please correct me if I am wrong or which other tools there are.
If it is daytrading that is less of an issue but still you rather would rate systems with the same profit in less bars higher than systems that need 2 times as much bars to have the same profit. Also, you could do more trades and could have less trades that go nowhere.
In SS I found that longer trades with the same profit definitely have a much worse OS/IS rate.

Quote: Originally posted by admin

Im still very open to this idea and its been passed onto the programer to comment, but here is an observation.
ts with wfo use to be able to optimize +=5% and +- 10% on the parameter range or both
of course this means opt takes 3 or 6 times longer.
It was many years since I did this, but I observed it degraded the OOS results.
I tried on quite a few systems.

admin - 7-4-2021 at 01:29 AM

@rws. no reply from me on this today. tied up on other things

Carl - 15-4-2021 at 01:47 AM

My livetrading CL strategy result has been flat for almost a year, but started to go up again!

admin - 15-4-2021 at 05:34 PM

Quote: Originally posted by admin

@rws. no reply from me on this today. tied up on other things

This is a job for the most senior programer, which means if it happens (Seriously being considered) it wont be for some months.
If one of the secondary programmers could do it, it gets done quickly

admin - 19-4-2021 at 01:51 AM

This is the new settings files for trial users.
They are much better than old settings and use the new filters of the latest GSB.
All users should look at these.
It has secondary filter closelesshighlow3 and filter closelessprevd (not normalized)
if your in beta tester mode, choose AIC, otherwise cross under entry types.

unzip data.zip into c:\gsb\data. Overwrite any files.
It has macros for es nq ym, data files etc
as is, you can see 50% approx out of sample using indicators I know work well on NQ
This is not how we build systems however. (see the most recent gold videos)

Note the following should be changed
change this to 100 , 0

choose nq/es/ym data. from optsettings
run macro 1 to get correct nq/es/ym settings
run macro2 to chose the top 10 inidcaotrs
run macro3 to build 50,000 systems and get the top 300 etc
Its recommend to run macro2&3 in GSB automation, but it doesn't have to be the case.

Attachment: Login to view the details

admin - 20-4-2021 at 04:30 AM

feedback on the above welcome. Good change it will work fine on some other markets too
Ive been out of action today. TS is constantly crashing when the markets open. This has kept me out of action for some time.

comment from user Carl
Just downloaded the 62.12 installer, did a clean install and ran macro1. Nothing happened.

It’s because GSB.exe.config is not included, macro1 sets the available entry modes to 0.

After running GSB.exe.config, 2 entry mode are activated and systems are being build.

I will get this fixed soon. It only affects trial users as long as you have gsb.exe.config (which gives you extra features and indicators)

Carl - 20-4-2021 at 05:36 AM

Quote: Originally posted by admin

feedback on the above welcome. Good change it will work fine on some other markets too
Ive been out of action today. TS is constantly crashing when the markets open. This has kept me out of action for some time.

comment from user Carl
Just downloaded the 62.12 installer, did a clean install and ran macro1. Nothing happened.
It’s because GSB.exe.config is not included, macro1 sets the available entry modes to 0.

After running GSB.exe.config, 2 entry mode are activated and systems are being build.

I will get this fixed soon. It only affects trial users as long as you have gsb.exe.config (which gives you extra features and indicators)

For now:
1. run macro1
2. check if one or more entry modes are enabled
If not, then manually enable entry mode cross in the left GSB menu.
3. start build process

Update:
I installed GSB again on another computer.
Everything is working fine. Entry mode cross enabled at start-up and after running macro1. Cannot reproduce the issue I saw a few days ago.

meldinman - 20-4-2021 at 08:42 AM

has anyone else been having a large amount of TS crashes this past week? Im wondering if there may be a bug in new GSB codes causing it?

NickW - 20-4-2021 at 09:29 AM

did you install new gsb scripts lately?
did you try to roll back to previous version?

not experiencing issues in MultiCharts.

meldinman - 20-4-2021 at 09:41 AM

Yes I im using new gsb scripts, which is why Im wondering if its a function of some code bugs. Working with TS now but seems like my system is getting tons of .net errors from TS charts

bartek - 20-4-2021 at 10:04 AM

Quote: Originally posted by admin

feedback on the above welcome. Good change it will work fine on some other markets too
Ive been out of action today. TS is constantly crashing when the markets open. This has kept me out of action for some time.

comment from user Carl
Just downloaded the 62.12 installer, did a clean install and ran macro1. Nothing happened.

It’s because GSB.exe.config is not included, macro1 sets the available entry modes to 0.

After running GSB.exe.config, 2 entry mode are activated and systems are being build.

I will get this fixed soon. It only affects trial users as long as you have gsb.exe.config (which gives you extra features and indicators)

For indicators finding do you use Auto nth day mode= All?

Carl - 20-4-2021 at 12:10 PM

Quote: Originally posted by bartek

For indicators finding do you use Auto nth day mode= All?

Personally I think using Nth day mode Trd or All are both okay.
As long as you don't use all the price data up to yesterday.
You have to keep 1 to 4 years of price data as out of sample data.

admin - 20-4-2021 at 05:38 PM

bartek
with inidcator testing, the systems are built with nth 80, see 80 trades in sample, 80 trades out of sample..
BUT they appear in the GUI as nth all. So this is 50% out of sample, plus there are years of data not seen yet as well

admin - 21-4-2021 at 05:16 AM

Quote: Originally posted by meldinman

has anyone else been having a large amount of TS crashes this past week? Im wondering if there may be a bug in new GSB codes causing it?

IM using the identical ts eld, but yes, I've had the worst week ever. I built a new trading computer. The old one was very stable but I'm upgrading it from 1 cpu 128gb to 2 cpu 384 gb ram
(Used for GSB after hours)

I don't think its gsb code though.
Its hard for gsb code to crash ts if you tried.
Im suspecting my crashes are due to a problem with TS trading app.
see screen shot. I saw it appearing intermittently in a box on the opt of ts.
This is just a wild guess, and I'm not saying its the issue.
corrupted workspaces, cache, windows registry issues etc are much more likely causes.
TS10 reputation is still poor. (I'm on 9.5 build 28)
Other things is you should reboot your pc once per week. TS is normally very stable as long as you don't open lots of charts going back to start of contract dates. This is almost certain to crash TS is my experience.
You can also go to add remove programs, and to a repair on TS
delete ts cache and tcache. Open all your charts, let the data load. wait for it to be loaded, then close charts, reopen charts.
Best to have min possible data on your charts. often 1 year is ok, but some systems like zone trader might need up to 3 years.
Its wise to backup your eld, and zip up your workspace. A zip file has crc (cyclic redundancy check) so any file coruptions would be clear.
you can also backup the entire ts folder using .7zip on max compression < 100 mb
if you delete cache and tcache.
also open just a few workspaces at a time, not 20 or so. (TS support told me that)

admin - 21-4-2021 at 05:22 AM

Quote: Originally posted by meldinman

Yes I im using new gsb scripts, which is why Im wondering if its a function of some code bugs. Working with TS now but seems like my system is getting tons of .net errors from TS charts

likely your .net is corupt. (just my guess)
Dont think I've ever had a .net ts error.
Corupt .net does affect gsb at times, but likely its a different version of .net to what GSB uses

admin - 21-4-2021 at 05:24 AM

Quote: Originally posted by admin

feedback on the above welcome. Good change it will work fine on some other markets too
Ive been out of action today. TS is constantly crashing when the markets open. This has kept me out of action for some time.

comment from user Carl
Just downloaded the 62.12 installer, did a clean install and ran macro1. Nothing happened.

It’s because GSB.exe.config is not included, macro1 sets the available entry modes to 0.
User Carl and I cant reproduce this error, so all seems fine

After running GSB.exe.config, 2 entry mode are activated and systems are being build.

I will get this fixed soon. It only affects trial users as long as you have gsb.exe.config (which gives you extra features and indicators)

meldinman - 21-4-2021 at 07:21 AM

ya, TS support is saying it mgiht be becuase im using Windows server 2019 edition, they say only support windows 10. Likely not a GSB issue, had to go through all the steps above, still seeing crashes with just the TS default strategies. Sometimes TS is the best but sometimes it makes me wanna cry

loclhero - 21-4-2021 at 01:17 PM

Trading System Interpretation

Does it make sense to interpret the weight associated with an indicator as a measure of the importance of that indicator?

admin - 21-4-2021 at 06:02 PM

Quote: Originally posted by meldinman

ya, TS support is saying it might be becuase im using Windows server 2019 edition, they say only support windows 10. Likely not a GSB issue, had to go through all the steps above, still seeing crashes with just the TS default strategies. Sometimes TS is the best but sometimes it makes me wanna cry

ts have said that for years about windows server, but I've been running on windows server for 5 or 10 years,and been stable.
I too felt like crying. Its taken up most of my time in the last week. Up to 3 am once.
are you on ts 9.5/28?
I got a second login as a test, and it seems stable, though maybe too early to tell.

someone posted about the ie message coming up in TS, which is annoying.
Im asking, why is TS accessing IE when I don't ask it to. They gave more complex fix and said put google into restricted sites.
I didn't do there recommendations as I was nervous about doing stuff with registry and dll midway in a trading day. I did block google and I think that fixed the issue of the script. Unlikely this will cause stability, but I'm not taking chances.
My suspicion is, its the tsapps that caused instability. Still my wild guess but my unstable ts has been better since I did this.
Very useful for me to know standard workspaces crash.
Im on my 3rd server to try and fix this and that didn't help.
You should just try a different machine and new operating system.
I can get you a VM 1 ms away from TS servers if you need to test it.

meldinman - 21-4-2021 at 06:20 PM

I told them I have a friend running on windows server no issues. The crashes started happening after I spent a long time migrating all my workspaces back to ts 9.5 from 10 in the hope of better stability. But instead of 1 to 2 crashes per week it started happening 4 to 5 times per day. Something in the install was giving TS hangups so I went after tech support to get to the bottom (not just their usual bs about rewriting workspaces etc)
My crashes where being generated by TS charting not TS apps. I dont have any apps running either and I think my server automatically blocked those connections. I never thought to try a different server, I'm using steadfast and they could probably get that done for me. What a nightmare either way, hope you clear up your issues as well.

admin - 21-4-2021 at 06:31 PM

Quote: Originally posted by meldinman

I told them I have a friend running on windows server no issues. The crashes started happening after I spent a long time migrating all my workspaces back to ts 9.5 from 10 in the hope of better stability. But instead of 1 to 2 crashes per week it started happening 4 to 5 times per day. Something in the install was giving TS hangups so I went after tech support to get to the bottom (not just their usual bs about rewriting workspaces etc)
My crashes where being generated by TS charting not TS apps. I dont have any apps running either and I think my server automatically blocked those connections. I never thought to try a different server, I'm using steadfast and they could probably get that done for me. What a nightmare either way, hope you clear up your issues as well.

i don't use any tsapps either, but I saw rectangular box up the center top pop up, and I suspect its looking to update them etc. But as I said I've deleted them and it seems better

meldinman - 21-4-2021 at 06:41 PM

how did you go about deleting them?

admin - 21-4-2021 at 07:33 PM

Quote: Originally posted by meldinman

how did you go about deleting them?

go to the el editor, open sort by strategy, then click the x delete button

jasonp - 22-4-2021 at 01:31 AM

I gave up with TS a few years back now for live trading however I still use their data. I connect the data to multi charts and use multicharts with Interactive Brokers. I run a portfolio of systems using MC portfolio trader.

TS 9.5 was still open and given me script errors so I upgraded to TS 10.0 and as I am not doing anything except using the data in TS 10 I have had no problems.

Multicharts has been really reliable. I leave it unattended for weeks at a time.

I would be really interested to hear how others are configuring their live trading?

admin - 22-4-2021 at 01:52 AM

Quote: Originally posted by jasonp

I gave up with TS a few years back now for live trading however I still use their data. I connect the data to multi charts and use multicharts with Interactive Brokers. I run a portfolio of systems using MC portfolio trader.

TS 9.5 was still open and given me script errors so I upgraded to TS 10.0 and as I am not doing anything except using the data in TS 10 I have had no problems.

Multicharts has been really reliable. I leave it unattended for weeks at a time.

I would be really interested to hear how others are configuring their live trading?

TS 9.5 is normally reliable if you reboot once a week, don't open lots of charts with data back to beginning of contracts.
I think MC is likely more stable than TS. IM not into aesthetics, but I find MC a dog to use.
The metrics I need to look at every trading system requires me to look at multiple tabs. Drives me nuts. You need all the info in one screen. I Wont use MC for reason alone.
On TS is on the performance summary. Its also more buggy as far as ts code goes, and lots of work went into getting mc to match ts. the code is 98% the same, and the 2% can be critical. TS is also much easier to use with its inbuilt data manager. MC is also a fraction slower in execution, regardless of the data feed. ie ts data feed iqfeed etc.
But this was about 4 years ago when I time-stamped orders to the microsecond

The issue is similar to do you want android or apple. Sometimes its a really easy choice. TS wont allow people in certain countries to have it, or ts doesn't have the data feed on a market you want to trade.

meldinman - 22-4-2021 at 06:18 AM

In addition to peters points, I use both MultiCharts through the IB gateway and TS. The Ib gateway logs itself out every weekend which is an utter Pita. Usually have to reset IB and MC to get it working again on mondays. It has its plusses but the way it goes about combining historical trades and real time trades is confusing and counterintuitive. Im not sure how others have it running for weeks on end but I usually have an issue with it every couple days where strategies aren't filling order or even sending them. Granted some of that might be due to IB gateway but still not as reliable as I would have hoped.

RandyT - 22-4-2021 at 07:08 AM

Guys, your discussion about TS issues jogs a memory for me. I've seen cases in the past where something about a workspace gets corrupted and causes constant crashes in TS. As I remember, the solution was to rebuild workspaces from scratch and the problem went away.

One of the many reasons that led me to move to Multicharts... (not that MC is perfect)

Carl - 22-4-2021 at 09:40 AM

Hi Randy,
Had the same issue a couple of times.
No error messages, but TS hanging constantly, really slow or not responding.

This is what I have been doing to solve issues, maybe these steps help other users as well.

I make a backup of my desktops and workspaces almost every day.

When TS issues start:
restart VM
restart TS
delete TS cache
windows start - settings - apps - Tradestation - change - next - repair
then finally load the back-up workspaces and desktops

admin - 22-4-2021 at 04:25 PM

Quote: Originally posted by meldinman

In addition to peters points, I use both MultiCharts through the IB gateway and TS. The Ib gateway logs itself out every weekend which is an utter Pita. Usually have to reset IB and MC to get it working again on mondays. It has its plusses but the way it goes about combining historical trades and real time trades is confusing and counterintuitive. Im not sure how others have it running for weeks on end but I usually have an issue with it every couple days where strategies aren't filling order or even sending them. Granted some of that might be due to IB gateway but still not as reliable as I would have hoped.

I have IBcontroler + some other scripts to keep the full tws open.
It logs out every day, then my scripts make it log back in < 1 minute later
Script will log you backin regardless of time of day that logout occurs
IBcontroler is free but the extra scripts are not in it.
For a small charge I could get this saleable. Requries me to document it furhter.

admin - 23-4-2021 at 05:49 AM

Quote: Originally posted by loclhero

Trading System Interpretation

Does it make sense to interpret the weight associated with an indicator as a measure of the importance of that indicator?

sorry I missed your comment.
Im not sure about this.
Reemeber all inidcators are normalized to be between -100 to + 100
If you use any indicator cross method (and there are pros and cons to this)
all indicators should have a weight of zero, so they are all equal.

if mutilpied together, a bigger weight means a smaller value needed to get an entry.
Regardless it doesn't matter what we the human thinks. the WF should choose the optimal parmater

meldinman - 23-4-2021 at 11:14 AM

I would be interested in something like that peter

admin - 23-4-2021 at 05:06 PM

Quote: Originally posted by meldinman

I would be interested in something like that peter

ok, give it a few days and I will see if there are other users who want it.
Was thinging $100 or less if there is more users.
It works really well, and better than just the gateway, as you can see your positions / do trades if needed in the gateway

bartek - 24-4-2021 at 03:54 PM

Using the newest macros Build NQ/ES/YM

Zrzut ekranu 2021-04-24 234958.png - 3kB

Zrzut ekranu 2021-04-24 235115.png - 21kB

Zrzut ekranu 2021-04-24 235157.png - 49kB

bartek - 26-4-2021 at 02:20 PM

Allowing pyramiding increased results.

Zrzut ekranu 2021-04-26 221850.png - 58kB

Zrzut ekranu 2021-04-26 221725.png - 6kB

ALiCa - 27-4-2021 at 06:57 AM

I understand that there is a lot learn before using the software...

I saw one of the last videos for CL, with the methodology explained.
From what I understand from the video it seems that Peter suggest to build the system with 3 time frames, 29, 30 and 31.
But also with 4 data series CL (Data1) HO, RB, NG as Data2,3 and 4.
I'm confused so we are not building "Against" 3 x 4 series... but we are building a system for CL on 3 and using the others symbols as secondary...
Am I right or I'm misunderstanding something?

RB, NG ad HO are "available" together with CL... so that the systems produced will require those series also in TS or Multicharts... does it make sense?

admin - 27-4-2021 at 05:58 PM

Quote: Originally posted by ALiCa

I understand that there is a lot learn before using the software...

I saw one of the last videos for CL, with the methodology explained.
From what I understand from the video it seems that Peter suggest to build the system with 3 time frames, 29, 30 and 31.
But also with 4 data series CL (Data1) HO, RB, NG as Data2,3 and 4.
I'm confused so we are not building "Against" 3 x 4 series... but we are building a system for CL on 3 and using the others symbols as secondary...
Am I right or I'm misunderstanding something?

RB, NG ad HO are "available" together with CL... so that the systems produced will require those series also in TS or Multicharts... does it make sense?

Whats good and bad is GSB methodology and GSB are under rapid development.
Whats bad is information gets absolute quickly.
Gold is the most recent material, and is best to follow. Also had great out of sample results. I no longer use mutil time frames ie 29, 30 31 at all in building or validating. Im also not using rb ho ng as data streams for validation.
whats important still on CL is not trading the first hour. A lot of work has been done for the last 2 months on CL, and there might be an update.
I think 15 /30 min CL is still the best data streams. (data1 cl15, data2 cl30
Gold videos are still really good, and there are just some minor changes I would make as updates, but they are not critical.
Next video will be on NQ/ES/YM
CL also has not been bullish from a systems point of view for some months, though that's improving of late. NQ ES right now I think are the hottest markets.

admin - 27-4-2021 at 06:32 PM

Quote: Originally posted by bartek

Allowing pyramiding increased results.

note you need "setstopshare;" added in the code to handle the stops properly.
next build of GSB will have this added.
I don't like to do more than 2 contracts, as otherwise you could for example get 3 entities on the same bar. Nice if you win and ugly if you get stoped.

ALiCa - 27-4-2021 at 07:29 PM

Question.

A quick clarification or any documentation about the difference between the 3 kind of filters:

1 - Indicators | We can enable all the available indicators as secondary filters
2 - SF Indicators | We can select Genetic Algo, CloseLess.... and others
3 - Filters | We can select IndicatorLevelCompare, AllowAllTrades. CloseLessPrevCloseD

I also understand that It's also possible to enable tertiary filters.... but my main confusion (and I'm not able to find any documentations is about the 3 filters above)
Many THANKS!

admin - 27-4-2021 at 07:45 PM

Quote: Originally posted by ALiCa

Question.

A quick clarification or any documentation about the difference between the 3 kind of filters:

1 - Indicators | We can enable all the available indicators as secondary filters
2 - SF Indicators | We can select Genetic Algo, CloseLess.... and others
3 - Filters | We can select IndicatorLevelCompare, AllowAllTrades. CloseLessPrevCloseD

I also understand that It's also possible to enable tertiary filters.... but my main confusion (and I'm not able to find any documentations is about the 3 filters above)
Many THANKS!

Excellent questions you have asked.
1) I think you mean as indicators, NOT as secondary filters. If so yes

2) You can, but there is no way you should so them all. You do one at a time and go through the entire methadolgly. This takes a lot of CPU time, requires GSB automation , and if you don't have the CPU time then let others do it. This is being done over time on various markets.
There is a number of users with decent server farms who work on things collectively, plus I also work with numerous volunteers on specific projects.

3) This works very well on many markets especially stock index. I have not tested for example on energies

4) As for Tertiray filters. They work and I have done NOTHING at all with them yet.
Im focusing on filters, and the closed filter works great on es/ym/nq using secondary filter hihghlowclose3. These settings & macors were posted in the forum a week or so ago for es nq ym

bartek - 27-4-2021 at 09:49 PM

Quote: Originally posted by admin

Quote: Originally posted by bartek

Allowing pyramiding increased results.

With this tip

admin - 27-4-2021 at 10:45 PM

@bartek. I dont understand what you wrote

bartek - 27-4-2021 at 10:52 PM

Quote: Originally posted by admin

@bartek. I dont understand what you wrote

Results increased even more with "setstopshare;" instruction

Zrzut ekranu 2021-04-28 065004.png - 15kB

Zrzut ekranu 2021-04-28 064902.png - 55kB

admin - 27-4-2021 at 10:57 PM

@barteck. thats because say you had $2000 stop, without setstopshare you have a $1000 stop when 2 contracts are open at the same time. Thats too tight.
worse of course when 3 contracts are used.

ALiCa - 28-4-2021 at 04:12 AM

Regarding...
1 - Indicators | We can enable all the available indicators as secondary filters
There is a switch to ENABLE each indicator as Secondary Filter, so I immagine that i can select Many and the Optimization will create populations that are also using the one I enabled as second filter...
I think it should be clear...

Regarding the
2 - SF Indicator | We can select Genetic Algo, CloseLess.... and others
Really I don't underatand Auto, Genetica Algo... What are they changing in the population?
Which is the intaraction they have with the filters I enabled on the built in indicators...

And Last... there is a third setting... Filters With 3 soptions only...
3 - Filters | We can select IndicatorLevelCompare, AllowAllTrades. CloseLessPrevCloseD
Same question... What are they changing...

Can you please send me a link to instructions, docs or a video... Or an explaination abotu the way they works...
It's really not easy to understand...
Grazie!

portfolioquanttrader2020 - 28-4-2021 at 07:26 AM

Hello, I am also a GSB user, if you want we can meet to learn. I have hired a person who is a translator.
My email is cammonam@gmail.com

meldinman - 28-4-2021 at 07:28 AM

1 - indicators are the first layer to your trading signal - example: enter long if price crosses above a moving average etc
2 - Secondary filters are a way to filter out signals - example: only take the moving average long in today's high is higher than yesterdays
3 - filters are a little more complicated and a newer feature. only take trades when 1 and 2 apply and 3 is also true.
Oversimplified but hope thats a helpful start. I would also suggest looking at the tradestation code generated by GSB to get a feel for GSB architecture.

admin - 28-4-2021 at 04:42 PM

Quote: Originally posted by ALiCa

Regarding...
1 - Indicators | We can enable all the available indicators as secondary filters
There is a switch to ENABLE each indicator as Secondary Filter, so I immagine that i can select Many and the Optimization will create populations that are also using the one I enabled as second filter...
I think it should be clear...

Regarding the
2 - SF Indicator | We can select Genetic Algo, CloseLess.... and others
Really I don't underatand Auto, Genetica Algo... What are they changing in the population?
Which is the intaraction they have with the filters I enabled on the built in indicators...

And Last... there is a third setting... Filters With 3 soptions only...
3 - Filters | We can select IndicatorLevelCompare, AllowAllTrades. CloseLessPrevCloseD
Same question... What are they changing...

Can you please send me a link to instructions, docs or a video... Or an explaination abotu the way they works...
It's really not easy to understand...
Grazie!

1) It is a very very bad idea to enable many secondary filters. This is covered in the more recent gold videos
The number one critical issue new GSB users made was to run the wrong secondary filter on a market. So for trial users they had only the option of auto. Ive had to undo that as now the best option for indices is the highlow3lessclose SF with the filter closedlessclosebpv
In auto it chooses the sf set under contracts.
ga uses whatever SF is chosen under indicators normalized
closelessClosedBPV sf is NOT normalized.
filter is NOT normalized. This allows you to combine a NOT normalized closed filter with a normalized SF.
These settings are critical to get the max out of GSB. They differ for each market.
inidices use filter closed with sf GA highlow3lessclose.
energies we used GA with a few possible SF. However I have not done work on energies since we had filters but I suspect not normalized filter is not good for energies

Filters is very new, not yet documented, and is a high potential area of growth in GSB
A lot more filters will be added. IE pattern filters, time masks etc

Sorry but there are many competing demands for development time, and only so much can be done in a day.

Your asking very good & important questions. These are the settings you must get right, and you need to follow the gold methadolgy to get the best out of GSB

admin - 29-4-2021 at 12:05 AM

There is a new thread here on the Dax and day of week filters
https://trademaid.info/forum/post.php?action=edit&fid=1&tid=...

admin - 30-4-2021 at 01:32 AM

Rather than do a video (for now) I'm doing how to build NQ/ES/YM systems in the help file.
Its much much faster for me to do a document than cut a video.
This material is 50% finished. Hope to have the rest done in a week.
https://trademaid.info/gsbhelp/BuildingNasdaqSP500orDowsyste...

portfolioquanttrader2020 - 30-4-2021 at 06:58 AM

I think you should put each test with its correct configuration and join a repository with the same name of the test where the macros and results are, because if not it is difficult to find out. Each test with its explanation and codes

admin - 30-4-2021 at 04:56 PM

Quote: Originally posted by portfolioquanttrader2020

I think you should put each test with its correct configuration and join a repository with the same name of the test where the macros and results are, because if not it is difficult to find out. Each test with its explanation and codes

agreed
see this post from today.
https://trademaid.info/forum/viewthread.php?tid=39&page=30
I will add to the file repository section as well a little later on

https://trademaid.info/forum/viewthread.php?tid=262#pid4902

bartek - 1-5-2021 at 04:54 AM

Quote: Originally posted by admin

Rather than do a video (for now) I'm doing how to build NQ/ES/YM systems in the help file.
Its much much faster for me to do a document than cut a video.
This material is 50% finished. Hope to have the rest done in a week.
https://trademaid.info/gsbhelp/BuildingNasdaqSP500orDowsyste...

Peter, that's great work you've done here.

portfolioquanttrader2020 - 1-5-2021 at 09:31 AM

Peter el trabajo que has hecho, es mucho mejor que los vídeos.
Agradecería que sigas esa línea de trabajo para nosotros, los clientes.
Te doy mi enhorabuena.
Gracias

Carl - 1-5-2021 at 10:04 AM

Quote: Originally posted by portfolioquanttrader2020

Peter el trabajo que has hecho, es mucho mejor que los vídeos.
Agradecería que sigas esa línea de trabajo para nosotros, los clientes.
Te doy mi enhorabuena.
Gracias

English please.

RandyT - 1-5-2021 at 10:10 AM

Easy to accommodate our non-english speakers

"Peter, the work you've done is much better than the videos.
I would appreciate it if you continue that line of work for us, the clients.
I give you my congratulations.
Thanks"

Carl - 1-5-2021 at 12:11 PM

Hi Randy,

How are you going with Python machine learning?

RandyT - 1-5-2021 at 12:28 PM

Quote: Originally posted by Carl

Hi Randy,

How are you going with Python machine learning?

Carl,

Made a small bit of progress and then have gotten distracted by a new server. I need to get back to this...

mdb - 1-5-2021 at 12:56 PM

Ive had to undo that as now the best option for indices is the

highlow3lessclose SF
with the filter closedlessclosebpv

Where are these two filters found, they don't seem to be listed in "manager" 1.0.62.05

portfolioquanttrader2020 - 1-5-2021 at 01:17 PM

I am doing the test of building systems with the Nasdaq, following the guide that you put in https://trademaid.info/gsbhelp/BuildingNasdaqSP500orDowsyste...
Doubts:
1. My second pass results are different. Why?
2. It seems to me that the results are not good, because the DD is similar to the profit. It is right?

portfolioquanttrader2020 - 1-5-2021 at 01:19 PM

I am doing the test of building systems with the Nasdaq, following the guide that you put in https://trademaid.info/gsbhelp/BuildingNasdaqSP500orDowsyste...
Doubts:
1. My second pass results are different. Why?
2. It seems to me that the results are not good, because the DD is similar to the profit. It is right?

portfolioquanttrader2020 - 1-5-2021 at 02:12 PM

Peter
Do gsb systems have indicators to locate the inputs and outputs of them?
How could I see those indicators that come in the code and plot them on the TradeStation Chart?

Carl - 1-5-2021 at 02:57 PM

Quote: Originally posted by mdb

Ive had to undo that as now the best option for indices is the

highlow3lessclose SF
with the filter closedlessclosebpv

Where are these two filters found, they don't seem to be listed in "manager" 1.0.62.05

Until the end of 2020 as secondary filter closelessclosedbpv was used.

Recently Peter added a series of new secondary filters to GSB. And closetohighlow3 works very well on stock indices.
Further improvement combined with tertiairy filter closedlessclosebpv.

So two different ways to build strategies on stock indices.
All these filters are included in GSB 62.05.

mdb - 1-5-2021 at 04:32 PM

Please see the pic, I don't see those filters listed here. Am I in the wrong place?

mdb - 1-5-2021 at 04:36 PM

I also have a question about building a "Price Data File" for GSB to use. For example using ES 1 minute bars, do I build the "data file" using 24 hours data, regular session time data, or just the 8:30 to 3:00 (central time) session that I will actually be trading?

Thanks, lots of stuff going on with GSB

Carl - 1-5-2021 at 11:48 PM

Quote: Originally posted by mdb

Please see the pic, I don't see those filters listed here. Am I in the wrong place?

Choosing the filters:

The secondary filter drop down menu shows a few different settings: Auto, geneticalgorithm and a few short cuts.

If you set this to "Auto" GSB uses the method that is set in the contract list. For ES: closelessprevclosedBPV not-normalized

If you set this drop down menu to geneticalgorithm, you can activate any indicator in "built-in indicators". These primary and secondary are normalized indicators. For example: stock market indices closetohighlow3.
Results can be further improved by adding a tertiary filter closedlessclosebpv.

Carl - 1-5-2021 at 11:56 PM

Quote: Originally posted by mdb

I also have a question about building a "Price Data File" for GSB to use. For example using ES 1 minute bars, do I build the "data file" using 24 hours data, regular session time data, or just the 8:30 to 3:00 (central time) session that I will actually be trading?

Thanks, lots of stuff going on with GSB

In the beginning a user needed to feed GSB with the data file you wanted to use. The problem was that when you wanted to try a different session, you had to go back to the trading platform (i.e. Tradestation) to export another data file.

GSB is now capable to build the data file you need by using an exported file from the platform.
The easiest is to export 1 minute regular session from your platform and save this to a txt file. Then construct the bars and session you need in GSB. You can construct any bar size and any session by using GSB.

But you can also use a ES 30 minute 0830-1500 data file straight from your platform.

There is no "right or wrong" here, just different methods to make it easy for the user. Works great!

admin - 2-5-2021 at 10:16 PM

Quote: Originally posted by mdb

Here they are

admin - 2-5-2021 at 10:25 PM

Quote: Originally posted by mdb

i would use 830 to 3pm @es.d
@es with session mask 830 to 1500 is 98% the same, but that's not 100%
the exception is if your going to enter before 830 etc, which is rare to do that sort of thing. Note however I have two 24 hour es systems that do better or exit after hours, but they are non GSB systems
see https://trademaid.info/gsbhelp/Systemsforsale.html
swingES, Overnight ES, and there likely be another swing ES which
has no relationship in trading to the current swingES

bartek - 3-5-2021 at 07:31 AM

From your NQ/ES/YM guide:

Quote:

5) Walk forward all members in this group. Choose systems that have got good walk forward results for live trading.

Maybe it is a dummy question but I am curious.
What is the reason to run WF for all family members as they differentiate only by parameters and we optimize just parameters with WF? Isn't sufficient to run WF only for one member?

admin - 3-5-2021 at 04:45 PM

Quote: Originally posted by bartek

From your NQ/ES/YM guide:

Quote:

5) Walk forward all members in this group. Choose systems that have got good walk forward results for live trading.

Good question.
You should expect somewhat different results in each wf.
The range used is relative to the initial parameters, and the WF is Genetic, so a random seed is also used in each WF.
Its also very easy to do, and not cpu or time intensive to run a number of WF.

Pages: 1 .. 6 7 8 9 10 .. 12