DislikedThanks for putting Spud's ideas into one concise document, nondisclosure00. I am sorry for your connection to the helicopter crash in Arizona.
Have you thought about posting your EA's on the MT Indicators and Experts group? There are some pretty talented programmers on that list as well as many eager testers willing to open another demo account for testing new ideas.Ignored
Well, If you can read 'rules', I need to figure out why the Spud EA's don't open any trades.
All the work is done in the stochOpen and stochClose procedures.
I don't think I've quite got the rules in those procedures to match those in Spud's methods.
Take a hack at it. Change it etc. It's there for everyone to play with.
I stumbled on this thread earlier today. I was introduced to Spud's MTF threads a couple of weeks ago; I really like this stuff. It's great to see someone sharing their knowledge with the community for free. So it was a great surprise to find out that someone is writing an EA for these systems. Great job, nondisclosure.
Anyways, here's my contribution to the EA. I took a look at the code for the SpudFrye Method II EA (based on the Escalator to Pips system), and I think I found the problem why it's not opening any trades.
The problem is in the 'bool stochOpen(string signal)' function. There are two problems here. They are both in the two if tests ( if(signal=="Short" && ...) and if(signal=="Long" && ...) ).
The first problem is that we are testing whether the stochastics are lining up in the 4 time frames using using the values for Stochastic(14,3,3) AND Stochastic(5,3,3). In other words, in order for a buy/sell signal to be generated the stochastic(14,3,3) for all time frames must be all going in the same direction AS WELL as the stochastic(5,3,3). So if either of these fail then no signal is generated. We should use either sto(14,3,3) or sto(5,3,3), but not both. At least this is my interpretation of SpudFrye's system.
Issue 2 is related to the following block of tests in the if statements (e.g. for the case where we're looking for a "Short" signal):
(main_15last > highlimit)
&& (main_30last > highlimit)
&& (main_1Hlast > highlimit)
&& (main_4Hlast > highlimit)
So let's break it down. The main_xxlast variables are the values of the stochastic for the next to last candle in the corresponding time frame. Keeping in mind that we are looking for a sell signal in this example, what we are testing in this block is for the next to last candle's stochastic value to be above the '80' line. For the buy signal it would be that the values are below the '20' stoch line. I understand why you are testing these, but I don't think it's necessary. What you are testing for is the ideal condition where the stochastic start off from the '80' line moving down in unison. I don't think we have to have the stochastic at the '80' line for all time frames. I think these 4 tests can be scrapped all together. Just look at the examples in the PDF instructions for this system and you will see that in none of the examples the 4 time frames are all starting from '80' (or '20') line.
There is one last little issue. I'm being picky on this one, as it is not really an error. In the rules it states that a trade is entered when all the stochastic lines are moving upwards [or downward for a sell] and the solid lines are all above OR AT 20 [or at 80 for a sell]. In other words if the lines are equal to 20 at that point then the signal is still good. So I would revise the test that reads 'main_xx < highlimit' [and the corresponding one for the long signal] to 'main_xx <= highlimit'.
Let me know if I did not explain clearly. I have not tested my theories yet, but I will as soon as I get a chance (probably tomorrow). I also have not looked at the rest of the code in depth for any more logic errors. Let me know what your thoughts are on this.
By the way, thanks for writing such clear code. It's very easy to read and follow. Kudos!
Hopefully we can all colaborate on this EA to get it working right.
- #25
- Edited 9:36am Aug 9, 2007 9:10am | Edited 9:36am
- Joined Apr 2007 | Status: Gettin' kick in the nutz every day! | 721 Posts
Hi nondisclosure,
I stumbled on this thread earlier today. I was introduced to Spud's MTF threads a couple of weeks ago; I really like this stuff. It's great to see someone sharing their knowledge with the community for free. So it was a great surprise to find out that someone is writing an EA for these systems. Great job, nondisclosure.
Anyways, here's my contribution to the EA. I took a look at the code for the SpudFrye Method II EA (based on the Escalator to Pips system), and I think I found the problem why it's not opening any trades.
The problem is in the 'bool stochOpen(string signal)' function. There are two problems here. They are both in the two if tests ( if(signal=="Short" && ...) and if(signal=="Long" && ...) ).
The first problem is that we are testing whether the stochastics are lining up in the 4 time frames using using the values for Stochastic(14,3,3) AND Stochastic(5,3,3). In other words, in order for a buy/sell signal to be generated the stochastic(14,3,3) for all time frames must be all going in the same direction AS WELL as the stochastic(5,3,3). So if either of these fail then no signal is generated. We should use either sto(14,3,3) or sto(5,3,3), but not both. At least this is my interpretation of SpudFrye's system.
Issue 2 is related to the following block of tests in the if statements (e.g. for the case where we're looking for a "Short" signal):
(main_15last > highlimit)
&& (main_30last > highlimit)
&& (main_1Hlast > highlimit)
&& (main_4Hlast > highlimit)
So let's break it down. The main_xxlast variables are the values of the stochastic for the next to last candle in the corresponding time frame. Keeping in mind that we are looking for a sell signal in this example, what we are testing in this block is for the next to last candle's stochastic value to be above the '80' line. For the buy signal it would be that the values are below the '20' stoch line. I understand why you are testing these, but I don't think it's necessary. What you are testing for is the ideal condition where the stochastic start off from the '80' line moving down in unison. I don't think we have to have the stochastic at the '80' line for all time frames. I think these 4 tests can be scrapped all together. Just look at the examples in the PDF instructions for this system and you will see that in none of the examples the 4 time frames are all starting from '80' (or '20') line.
There is one last little issue. I'm being picky on this one, as it is not really an error. In the rules it states that a trade is entered when all the stochastic lines are moving upwards [or downward for a sell] and the solid lines are all above OR AT 20 [or at 80 for a sell]. In other words if the lines are equal to 20 at that point then the signal is still good. So I would revise the test that reads 'main_xx < highlimit' [and the corresponding one for the long signal] to 'main_xx <= highlimit'.
Let me know if I did not explain clearly. I have not tested my theories yet, but I will as soon as I get a chance (probably tomorrow). I also have not looked at the rest of the code in depth for any more logic errors. Let me know what your thoughts are on this.
By the way, thanks for writing such clear code. It's very easy to read and follow. Kudos!
Hopefully we can all colaborate on this EA to get it working right.
I see what your saying above. I'll try to get to making changes to this next week. I've got a lot going on here at home I need to take care of.
Please, feel free to edit code, test, post, re-test, etc. The EA is for everyone! Just add your name to the top in the comment block at the top to get some credit!
ps. please, feel free to look at the rest of the code too!
EA First method, please correct me or add any missed details.
Entry BUY when :
- H4 stoch move up from low limit , the last value > low limit, this is the one switch only (if there is no Stoch H4 this way, there is no buy)
- H1 direction is up and below level 60
- M30 direction is up and below level 60
- M15 below 50 any direction
- Buy entry is wait for the next 3 hour after the H4 stoch was given
- if there is H1,M30,M15 condition above the buy signal is expired
- ...
rgds
no need for you to work on my code cuz i got some bullet proof ideas last night(well i hope) so im gonna change the code very much. but if you got the time i got a question aboute my code...
have succeded to get information from 4 timeframes whith my code ?
if you have the time to look at it i would appriciate it.
thx for the EAs you have posted, they have helped me alot.
damn i need to lern english
have succeded to get information from 4 timeframes whith my code ?
if you have the time to look at it i would appriciate it.
thx for the EAs you have posted, they have helped me alot.
damn i need to lern english
- #34
- Aug 18, 2007 1:39pm Aug 18, 2007 1:39pm
- Joined Apr 2007 | Status: Gettin' kick in the nutz every day! | 721 Posts
no need for you to work on my code cuz i got some bullet proof ideas last night(well i hope) so im gonna change the code very much. but if you got the time i got a question aboute my code...
have succeded to get information from 4 timeframes whith my code ?
if you have the time to look at it i would appriciate it.
thx for the EAs you have posted, they have helped me alot.
damn i need to lern english
