EA Factory

Quoting GirlFlyer

Disliked

Calling all EA Programmers!

I have recently discovered a problem with MT4's Strategy Tester and my EA and I was wondering if anybody else was noticing this problem or if it is a known problem and what can be done about it.

When my EA executes with Strategy Tester, it is pulling the data out of the indicators properly at the right time etc. However, I discovered that the prices for the Bid/Ask are coming from a time frame 2 hours earlier than the time frame of the most current bar on the chart. I am at a loss to understand why this is occuring. I'm thinking it has something to do with a GMT time adjustment, but it shouldn't be happening. Anyway, it is throwing off all the entry prices on the EA completely and invalidating all my tests.

Help would be appreciated.

GirlFlyer

Ignored

Quoting aha

Disliked

GirlFlyer,

I haven't met that before. Could you post your EA here and describe what you did so that we can reproduce the problem?

aha

Ignored

Quoting GirlFlyer

Disliked

Aha,

Attached is a chart of a trade at the opening. The arrow represents the time and the price at which the trade is purchased at. However, you will notice that the price shows at the 1.9085 level (the computer puts these arrows at the purchase price). The trade is a valid trade according to my rules and the chart bar. But, the bar during that timeframe shows no valid price range that includes the price purchased at. I'm kind of at a loss on this problem. When I actually ran the EA, I had it print the values in the journal, and yep sure enough, its pulling it from 2 hours earlier....

Thanks for the help.

Ignored

Quoting aha

Disliked

Hi guys,

I think things are going a little too fast here. I mean I am very glad to see so many talented programmers focusing on TD SEQ indicator development -- that's enviable resource that can lead to something great and I will exploit that with you later -- But instead of working as a team, we are going all different directions, making it more like a competition than cooperation, and without even achieve a stable version yet. I don't think this is a health environment and will also quadruple our tester's effort too.

My proposal: Let's focused on making a stable version first, which I believe we are pretty close now. Then based on that we discuss a set of features we want to add, then either working together to code it or let one programmer code it and others review the source code.

In order to avoid confusing this thread even more, I will stop developing TD Combo based on my TD Seq, and will offer to share my source code with Clam, GP2X, and Hans.

I suggest that our testers focus the testing effort on comparing GP2X's version with mine, which seem to incorporate the most complete set of logic behind TD Seq so far (correct me if I am wrong). Our programmer will inspect the structure of all 4 versions and vote on which one has the best structure to work on. Based on this one, we will implement the complete set of logic and features and release that as our first EA factory product.

Let's pool our talents together, how does that sound?

aha

Ignored

Quoting aha

Disliked

Hi guys,

I think things are going a little too fast here. I mean I am very glad to see so many talented programmers focusing on TD SEQ indicator development -- that's enviable resource that can lead to something great and I will exploit that with you later -- But instead of working as a team, we are going all different directions, making it more like a competition than cooperation, and without even achieve a stable version yet. I don't think this is a health environment and will also quadruple our tester's effort too.

My proposal: Let's focused on making a stable version first, which I believe we are pretty close now. Then based on that we discuss a set of features we want to add, then either working together to code it or let one programmer code it and others review the source code.

In order to avoid confusing this thread even more, I will stop developing TD Combo based on my TD Seq, and will offer to share my source code with Clam, GP2X, and Hans.

I suggest that our testers focus the testing effort on comparing GP2X's version with mine, which seem to incorporate the most complete set of logic behind TD Seq so far (correct me if I am wrong). Our programmer will inspect the structure of all 4 versions and vote on which one has the best structure to work on. Based on this one, we will implement the complete set of logic and features and release that as our first EA factory product.

Let's pool our talents together, how does that sound?

aha

Ignored

Quoting GirlFlyer

Disliked

Aha,

Nevermind.... I just figured out the problem. There is a 2 hour difference between my 1 minute charts and my 15 minute charts.... seems as if Alpari does not use the same time index as IBFX.

A little irritating...

GF

Ignored

Quoting clam61

Disliked

i was always going to share my code, but then you didn't so i thought "hmm maybe i shouldn't either", haha.

im open to give my code to anybody, but its a bit messy right now. overall its pretty simple code and straight forward, just one loop to calculate the #s and one loop to display the numbers.

unfortunately i dont think we would be able to share bits of code and build upon each other, given that this is not object oriented software by any means. most likely each of us has priority helper functions that cannot be reused.

with that in mind, we should choose one person's code and perhaps build upon that.

before we decide whose code to build upon, lets list off what our indicators can do and then we can build off the one with the most features (no sense redoing stuff).

This is the summary of my version:

1. Configurable colors
2. Horizontal lines box the true high and true low of the setup period
3. You have the option to display the #s and arrows or not. This can help reduce clutter
4. An arrow is displayed on a 9 setup
5. A P is displayed on a perfect 9 setup
6. An arrow is displayed on a 13 setup
7. A P is displayed on a perfect 13 setup
8. A different style arrow is displayed on a 2 bar flip setup
9. Another different style arrow is displayed on a 4 bar flip setup
10. A stop loss line is drawn based on the setup and countdown period
11. An "x" is draw when the setup or countdown is nullified

Ignored

Quoting GP2X

Disliked

Agree, I will stop developing mine until we all agree on the logics behind. For the visual features, it might not be so important if our final goal is an EA.
Another way might be to have each one focus on one strategy, I still remember the posts about MurreyMath system and it seems worth to try. I also found some trend line documents, maybe we can use them to continue optimizing the mouteki EA.

Ignored

Quoting clam61

Disliked

i will edit my code then post it. anybody who wants to run with it can do with it what they want.

btw, is there any charting software that can allow you too look at stocks that also lets you make custom indicators? the only one i know of is TradeStation, but that one is expensive.

Ignored

Quoting clam61

Disliked

hi hans

yeah that is good, but i want to be able to display any stock i want, and not just the ones they allow

Ignored

Quoting aha

Disliked

One on each strategy is probably better than working together on one strategy. As for TD-Seq, if we get the logic right, I don't mind stand aside and let one of you finish it in your way. We have more than enough to work on.

Ignored

Quoting GP2X

Disliked

So which one do you perfer? Is your mouteki EA mature, do you want to continue that? I am wondering if we can try the combinations, for example, we can use the TD setup as a filter for trend line break. Only enter if the break happens after a perfrect TD setup. Or combine it with Vegas fib levels, enter when the setup or coutdown ends at fib levels. I wrote a Vegas 1 hour EA, which is profitable in backward test for EURUSD pair, the exit is perfect but the entry is not, too many whipsaws. Maybe I can use TD setup to replace the entry of it. There are too many things to try, just want to make sure we don't do the same again.

Ignored

Quoting GP2X

Disliked

So which one do you perfer? Is your mouteki EA mature, do you want to continue that? I am wondering if we can try the combinations, for example, we can use the TD setup as a filter for trend line break. Only enter if the break happens after a perfrect TD setup. Or combine it with Vegas fib levels, enter when the setup or coutdown ends at fib levels. I wrote a Vegas 1 hour EA, which is profitable in backward test for EURUSD pair, the exit is perfect but the entry is not, too many whipsaws. Maybe I can use TD setup to replace the entry of it. There are too many things to try, just want to make sure we don't do the same again.

Ignored

Quoting hans_gregor

Disliked

well, these servers have only 27-45 stocks

Ignored