| Pages:
1
..
40
41
42
43
44
..
98 |
tornado
Junior Member
Posts: 9
Registered: 15-6-2020
Member Is Offline
|
|
Actually i am running d:\GSB\Workers . i have sent teamviewer.com details .
Thanks。
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
Happy Saturday all,
Question about the difference between forcing use of an indicator vs using it as a secondary filter. What is the difference in these two scenarios?
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by bizgozcd  | Happy Saturday all,
Question about the difference between forcing use of an indicator vs using it as a secondary filter. What is the difference in these two scenarios?
|
if you had say 44 primary indcators, but are using # of indicators as 3
and you force rsi as true,
then rsi will be in every system and the other 2 indicators will be made up of the
ones from the 44
This can be done for SF too. Note I never use the force feature myself.
Not sure if this reply helps fully answer what your after. Ask again if needed
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
| Quote: Originally posted by admin | | Quote: Originally posted by bizgozcd | Happy Saturday all,
Question about the difference between forcing use of an indicator vs using it as a secondary filter. What is the difference in these two scenarios?
|
if you had say 44 primary indcators, but are using # of indicators as 3
and you force rsi as true,
then rsi will be in every system and the other 2 indicators will be made up of the
ones from the 44
This can be done for SF too. Note I never use the force feature myself.
Not sure if this reply helps fully answer what your after. Ask again if needed |
What I mean to ask is, if I use an indicator on a given system either as SF or Forced, all other parameters the same, would the results theoretically
be the same?
Thanks Peter.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by bizgozcd | | Quote: Originally posted by admin | | Quote: Originally posted by bizgozcd | Happy Saturday all,
Question about the difference between forcing use of an indicator vs using it as a secondary filter. What is the difference in these two scenarios?
|
if you had say 44 primary indcators, but are using # of indicators as 3
and you force rsi as true,
then rsi will be in every system and the other 2 indicators will be made up of the
ones from the 44
This can be done for SF too. Note I never use the force feature myself.
Not sure if this reply helps fully answer what your after. Ask again if needed |
What I mean to ask is, if I use an indicator on a given system either as SF or Forced, all other parameters the same, would the results theoretically
be the same?
Thanks Peter. |
if all indicators and all parameters were the same, then results would be the same.
Hope this answered the question ok
Thanks received (1):
+1 bizgozcd at 2020-08-11 09:30:52
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Peter, i was going to test tertiary filters, (since i remember i read somewhere that you enabled that?) however it seems that they are not enabled in
.620 and it seems that in statistics its not possible to perform same kind of test as for SF, to be able to see what (in this case) tertiary filters
is being used in top n systems..
Am i mistaken that tertiary filters shall be enabled in this release?
Best trading day using GSB systems ever for me yesterday in SI, congrats to those that was trading that market yesterday..
11.5k usd per contract..
Regards
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Daniel UK1 | Peter, i was going to test tertiary filters, (since i remember i read somewhere that you enabled that?) however it seems that they are not enabled in
.620 and it seems that in statistics its not possible to perform same kind of test as for SF, to be able to see what (in this case) tertiary filters
is being used in top n systems..
Am i mistaken that tertiary filters shall be enabled in this release?
Best trading day using GSB systems ever for me yesterday in SI, congrats to those that was trading that market yesterday..
11.5k usd per contract..
Regards |
Not sure how you did that on SI. Perhaps you have different session time.
I got 7775 per contract
tf are working but disabled. I wont enable till we have the ts code and stats working. Im very keen to get them, but its in the hands of the primary
programer
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Hi Peter perhaps you are using the small contract? which multiplier are you using?
I am using multiplier 5k
So my number is correct i believe, unless IB has an error 
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Daniel UK1 | Hi Peter perhaps you are using the small contract? which multiplier are you using?
I am using multiplier 5k
So my number is correct i believe, unless IB has an error
|
Im on the 5k, but i think our times are different. my times are central.
session time on ts (exchange time) is 8am to 1330
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Ahh, ok ok, i am on eastern, and my session time for SI used is 800 to 1430.
Anyway, was a very good trade, Congrats.. 
|
|
|
eurtrader
Junior Member
Posts: 3
Registered: 6-4-2019
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Daniel UK1 | Peter, i was going to test tertiary filters, (since i remember i read somewhere that you enabled that?) however it seems that they are not enabled in
.620 and it seems that in statistics its not possible to perform same kind of test as for SF, to be able to see what (in this case) tertiary filters
is being used in top n systems..
Am i mistaken that tertiary filters shall be enabled in this release?
Best trading day using GSB systems ever for me yesterday in SI, congrats to those that was trading that market yesterday..
11.5k usd per contract..
Regards |
Hi Daniel,
Is this system on SI your own system, or can I find the code somewhere on the GSB forum? Thanks!
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
| Quote: Originally posted by eurtrader | | Quote: Originally posted by Daniel UK1 | Peter, i was going to test tertiary filters, (since i remember i read somewhere that you enabled that?) however it seems that they are not enabled in
.620 and it seems that in statistics its not possible to perform same kind of test as for SF, to be able to see what (in this case) tertiary filters
is being used in top n systems..
Am i mistaken that tertiary filters shall be enabled in this release?
Best trading day using GSB systems ever for me yesterday in SI, congrats to those that was trading that market yesterday..
11.5k usd per contract..
Regards |
Hi Daniel,
Is this system on SI your own system, or can I find the code somewhere on the GSB forum? Thanks! |
Hi,
Its my own system that i developed.
You can find some very good systems that Peter have kindly shared in the private forum for CL and ES.
Cheers
|
|
|
tornado
Junior Member
Posts: 9
Registered: 15-6-2020
Member Is Offline
|
|
Hi Peter ，
I watched some videos and forums on how to build the system. Regarding Left side Features-Optimization, I saw different settings.
1. Training%-100, Test%-0, Validation%-0,
2. Training%-40, Test%-60, Validation%-0,
3. Training%-40, Test%-30, Validation%-30,
When building the system, which one should I choose, I am confused.
What is the difference between test and validation? Are they all out-of-sample data?
Thanks in advance.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by tornado | Hi Peter ，
I watched some videos and forums on how to build the system. Regarding Left side Features-Optimization, I saw different settings.
1. Training%-100, Test%-0, Validation%-0,
2. Training%-40, Test%-60, Validation%-0,
3. Training%-40, Test%-30, Validation%-30,
When building the system, which one should I choose, I am confused.
What is the difference between test and validation? Are they all out-of-sample data?
Thanks in advance. |
If your having a quick play with GSB, fine to use say 30% in sample and see what comes out.
However the methodology is using 100% training, but has out of sample defined by dates field. (typcailly 2015 or 2018) Test and validation are both
out of sample periods, but when you the human look at them and make decisions on results, they arnt really out of sample then
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
Hello again.
Question about trading periods. If I set trading periods to 10-12 but I am using MOC for a session that is 9-4pm, will the MOC be executed at 4pm?
Thank you.
|
|
|
RandyT
Member
Posts: 123
Registered: 5-12-2019
Location: Colorado, USA
Member Is Offline
|
|
| Quote: Originally posted by bizgozcd | Hello again.
Question about trading periods. If I set trading periods to 10-12 but I am using MOC for a session that is 9-4pm, will the MOC be executed at 4pm?
Thank you. |
Correct.
And note that a 4:00 bar closes at 4:00, so if the market closes at 4:00, your order won't be filled until the market reopens. If this is the case,
then a session time close at 3:59 will actually close your trade while the market is still open.
Thanks received (1):
+1 bizgozcd at 2020-08-17 15:51:38
|
|
|
Gregorian
Junior Member
Posts: 97
Registered: 23-5-2017
Member Is Offline
Mood: No Mood
|
|
Forcing minimum Average Trade in WF
Is it possible to force the WF process to produce only a result with an Average Trade at or above a certain level? Similar to how Performance Filter -
Full Period works. I tried setting Performance Filter - Walk Forward Min. Avg. Trade, but I still get a WF Current result with an Average Trade below
the specified minimum value.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by Gregorian | | Is it possible to force the WF process to produce only a result with an Average Trade at or above a certain level? Similar to how Performance Filter -
Full Period works. I tried setting Performance Filter - Walk Forward Min. Avg. Trade, but I still get a WF Current result with an Average Trade below
the specified minimum value. |
I will ask programer about this. It may be that no option with the higher at was avilable.
you can also tweak fitness to have a high at bias, but if fitness is np*at this is a hihg at bias. np*at^2 for example would be higher at still. I
dont think its a good idea (np*at^2) as number of trades will be too low, and then its not a robust sample
|
|
|
bizgozcd
Junior Member
Posts: 53
Registered: 27-5-2020
Member Is Offline
|
|
I'm having an issue that I've had before intermittently when choosing data, but I don't know how I resolved it last time. I thought I was making an
error in which data I was selecting, but the issue seems to be with regard to selecting data. For some reason, in the data selection window, no matter
what data I choose, GSB reverts back to the data streams I had previously selected and rejects the new selection with an Invalid Price Data error from
a previous selection.
I've tried resetting the price data cache multiple times, restarted GSB multiple times, and I used the "deselect" button in the data selection box
each time.
I've been stuck and unable to use GSB as a result, so if anyone has any ideas, please let me know ASAP. Thank you.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by bizgozcd | I'm having an issue that I've had before intermittently when choosing data, but I don't know how I resolved it last time. I thought I was making an
error in which data I was selecting, but the issue seems to be with regard to selecting data. For some reason, in the data selection window, no matter
what data I choose, GSB reverts back to the data streams I had previously selected and rejects the new selection with an Invalid Price Data error from
a previous selection.
I've tried resetting the price data cache multiple times, restarted GSB multiple times, and I used the "deselect" button in the data selection box
each time.
I've been stuck and unable to use GSB as a result, so if anyone has any ideas, please let me know ASAP. Thank you.
|
I will be happy to help via remote support. Very hard to fix this by message, and really simple by remote support. email me your details
peterzwag@gmail.com
Thanks received (1):
+1 bizgozcd at 2020-08-25 09:34:35
|
|
|
AlwaysLearning
Junior Member
Posts: 11
Registered: 22-3-2020
Member Is Offline
|
|
Hello,
I've just noticed an issue I don't understand yet, where a swing trading position, that did not occur in realtime, suddenly appears in mid-trade when
the tradestation platform reloads.
I only noticed it now, but the best detail I can provide to describe the situation is this:
I'm following a swing trade system on ES in realtime, and in realtime I took the trade that ended in the profit target on Aug 26th at 1430 - but then
no new trade opened ever since in realtime, however then on Aug 27th at 7:48pm, I reloaded the chart, and then a new trade appeared retroactively in
the history as though it opened on 2:30pm on Aug 26th a day earlier where in realtime no such trade existed
I believe the cause to be that the data feed may have changed by a little between realtime and historical reload causing a trade to trigger,
But if I want to follow these trades realtime, it means now I have to decide whether to join this trade mid-position at whatever price I get, or
decide not to take it
It makes me question how I might go about handling this situation, and whether anyone else has a solution or advice for it. Should I be reloading my
tradestation charts a few times a day to minimize this, or what can I do to learn more about this?
Thanks
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
| Quote: Originally posted by AlwaysLearning | Hello,
I've just noticed an issue I don't understand yet, where a swing trading position, that did not occur in realtime, suddenly appears in mid-trade when
the tradestation platform reloads.
I only noticed it now, but the best detail I can provide to describe the situation is this:
I'm following a swing trade system on ES in realtime, and in realtime I took the trade that ended in the profit target on Aug 26th at 1430 - but then
no new trade opened ever since in realtime, however then on Aug 27th at 7:48pm, I reloaded the chart, and then a new trade appeared retroactively in
the history as though it opened on 2:30pm on Aug 26th a day earlier where in realtime no such trade existed
I believe the cause to be that the data feed may have changed by a little between realtime and historical reload causing a trade to trigger,
But if I want to follow these trades realtime, it means now I have to decide whether to join this trade mid-position at whatever price I get, or
decide not to take it
It makes me question how I might go about handling this situation, and whether anyone else has a solution or advice for it. Should I be reloading my
tradestation charts a few times a day to minimize this, or what can I do to learn more about this?
Thanks
|
I wouldnt reenter the trade, as your too much in the money. If it goes down to your price level, it could also be a loosing trade. But its painful
regardless as you might miss a decent trade
Some people delete ts cache one a week. Im not decided if this is a good idea.
If you do, then get one chart to pull in all your data (with no strategies on it)
,close that chart, then open your trading charts.
I used to run 3 ts, to compare trades, but thats expensive for data feeds if your professional status. Turning off a chart or reloading during market
close is too problematic to me, plus im asleep.
|
|
|
AlwaysLearning
Junior Member
Posts: 11
Registered: 22-3-2020
Member Is Offline
|
|
| Quote: Originally posted by admin | | Quote: Originally posted by AlwaysLearning | Hello,
I've just noticed an issue I don't understand yet, where a swing trading position, that did not occur in realtime, suddenly appears in mid-trade when
the tradestation platform reloads.
I only noticed it now, but the best detail I can provide to describe the situation is this:
I'm following a swing trade system on ES in realtime, and in realtime I took the trade that ended in the profit target on Aug 26th at 1430 - but then
no new trade opened ever since in realtime, however then on Aug 27th at 7:48pm, I reloaded the chart, and then a new trade appeared retroactively in
the history as though it opened on 2:30pm on Aug 26th a day earlier where in realtime no such trade existed
I believe the cause to be that the data feed may have changed by a little between realtime and historical reload causing a trade to trigger,
But if I want to follow these trades realtime, it means now I have to decide whether to join this trade mid-position at whatever price I get, or
decide not to take it
It makes me question how I might go about handling this situation, and whether anyone else has a solution or advice for it. Should I be reloading my
tradestation charts a few times a day to minimize this, or what can I do to learn more about this?
Thanks
|
I wouldnt reenter the trade, as your too much in the money. If it goes down to your price level, it could also be a loosing trade. But its painful
regardless as you might miss a decent trade
Some people delete ts cache one a week. Im not decided if this is a good idea.
If you do, then get one chart to pull in all your data (with no strategies on it)
,close that chart, then open your trading charts.
I used to run 3 ts, to compare trades, but thats expensive for data feeds if your professional status. Turning off a chart or reloading during market
close is too problematic to me, plus im asleep.
|
Thank you for providing these helpful points about all these considerations
What I think I'll do now is track how many of the trades this occurs to, and I'll experiment with the ts cache as you describe - thanks for the
guidance on that
Your answers are helpful enough already, but one last thing I hope to ask is:
Is this kind of historical trade appearance commonly experienced by people, and something to account for in my trading plan, or is this more rare?
I'm trying to tell if this is something I would work attentively to keep on top of it, or if it's something I can just let happen and accept as
irreducible data error, and just be prepared to see trades appear that I'm to ignore if they show up like this.
And is there anything like a ballpark estimate of how much this affects the expected returns from backtests? -- I am imagining it just falls under the
same amount of trade variance we normally are to expect but I'm making sure to ask just because I don't want to make preventable mistakes in this new
area.
Thank you
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Im not sure. There are two issues.
Trades that occur that shouldnt, and trades that should occur and dont
you could always at market close run ts on a second computer and verify the trades, or get 2 or 3 ts logins
Peter
Thanks received (1):
+1 AlwaysLearning at 2020-08-27 20:49:36
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Hi Always learning.... for me this (ltrades that according to backtest should be there is not here or vice versa) happens very rarely, i use MC....
in the cases it has happened... it has been between rolls.. i switch between custom future and continues, and this is for CL and NG where one has to
roll before data provider rolls contract, due to new IB rules for energies... anyway... for it happens due to this and cause ar known, i beliebe
issue is when cause is NOT known, as in your case i believe... I often think that one trade out of 1000 does not really matter... operational issues
like this is hard to avoid to 100%.
Cheers
Thanks received (1):
+1 AlwaysLearning at 2020-08-28 16:01:57
|
|
|
| Pages:
1
..
40
41
42
43
44
..
98 |
![[*]](images/davis/default_icon.gif){kind=icon}
