Posts made by MrDaisyBates

I am pretty comfortable with FxDreema, but I'm no coder. Just as a heads up.

My aim/direction I'm going is building a centralized EA for the purposes of higher timeframe market conditions. It could either be an EA or a script (I'm indifferent at this second).

I would like this higher timeframe EA to basically output various price levels and directional bias. I would like to store that information into a place that is safe and not lost if the terminal needs to be restarted, etc.

Then, I would like my other EAs to run on lower timeframes and basically be referencing the values/information from the biasing one.

So, for example, the low timeframe EAs would basically be entry models with various target and stop management, but a lot of the relevant information would come from another EA that's responsible for pre filtering conditions and all of this.

Furthermore, I could see myself having a high timeframe algo that keeps all of this information updated, but then possibly have a script that I can use to populate the data for the first time (or any time I need to start from scratch).

So, for example, on first use, I would run the script, it would scan (loop through) how many ever candles necessary to populate an unknowably sized list of variables, and then the EA would just keep that information updated (adding and removing values as necessary). And, again, the lower timeframe EAs would just tap into this information.

I can't say much more because I haven't hammered it out yet.

I can tell you right now, I don't know how to use/code arrays. I figure they're PROBABLY required here somewhere. I really am 100% code illiterate.

Furthermore, I don't know if I need to be saving these into a file and then referencing them across other things OR if it's better to use the "Global Variables" that MT5 offers? (I've never used those, either).

So, what am I looking for/asking?

"What does this look like and entail exactly? And what skills/knowledge do I need to learn/master to be able to execute this way?"

I would also appreciate either tutorial/walkthroughs (very thorough because I'm dense ahahah) or at least referrals where you'd recommend me learn said skills. HOWEVER, I understand that's my problem and not yours. So, if you at least could point me in the right direction, I would really humbly appreciate it and anything above that is just even more appreciated ahaha.

Thanks in advance,

Thanks for the post. I know it's old, but if you were using MT5 instead of MT4, is this a more feasible route than using the "Global Variables" in MT5? Are there pros and cons to weighing each approach? I've never tried either, but a lot of criteria I like to use relies on variables and it's super aggravating when they're lost due to resetting something.

@TipsyWisdom I appreciate you sharing this. It's probably outside of Dreema's capability to ensure only one instance of it, though? Like, if you distributed it to a person that they're not sharing it with 10 other people?

Would it easily be able to be added where like...the creator could ensure only one instance of it can be used?

I assume a web service could be used to poll it somehow or whatever.

Could the account number or something (if the person will share it with you) be hardcoded in, too, where it would reject an account that wasn't a match? Or is there maybe a simpler/easier way than this that I don't know?

An example is also appreciated, but not required. I'd be more concerned instead of as a business model, if I like shared one with a friend or something not wanting anyone else to be able to benefit from it.

Thanks again for your contribution!

@l-andorrà by chance, does this block (the current market one) (as well as the current timeframe one) work in the strategy tester for MT5?

Leaving this here in case it happens to you!

If you reload the project (go to load project), it'll reload them on the screen.

And if you've never had this issue, dragging the little selection box with the mouse does NOT go past the edge of the pane with the node options listed (and the node I'm trying to select is right behind it)

Hey guys and gals,

I had selected a bulk group of nodes, dragged them to the left, some of them went off the edge of the screen, then I selected another node (deselcting the group). Now I have nodes off the left side of the screen and I can't find any way to select them, scroll over, etc.

Advice? Did I break the project?

SOLUTION (If it happens to you): Reload the project. Hit the "load project" button up there and select the project. It will move the nodes back on screen.

@roar Thanks, Roar! Hahaha. You're the best.

Surprisingly, I did NOT expect it to be that complicated! hahaha. I'm still a complete nub with custom coding (hence why I'm drawn to FxDreema). I don't fully grasp it still, but I think you've given me more than enough to try and figure it out. Thanks, bud!

@bk7 Ah! That's an excellent work around that I hadn't considered! Thanks for that. But, like you said, that weekend gap may cause issues and this is counting hourly candles, so that may be an issue (but it may be one I can work around, so I'll surely keep this in mind!)

I'll try to keep this brief and to the point. I'm coding for MT5 if it's relevant, but I don't think it is.

I have an EA that basically runs a loop and if it identifies a candle with a specific condition, I take the candleID from it, and I also store the time for the candle in another variable.

Here's what I need it to do:

Once per bar, I want it to check if the candle that happened at that time stored is more than a certain amount of bars away, that way I can have it reset something.

Here's a better example if you're having a difficult time visualizing it.

On init, it scans and finds that candle id 10 meets the condition. The candle id is stored and the time is also stored. Once per bar, I want it to check if that candle is more than 24 bars away, for example, that way I don't have to run the ENTIRE loop once per bar to see how far back the event occurred.

The loop was fine, the storing of those variables is fine. I cannot find a node/condition/etc that allows me to look up a candle ID by its time so that I can somehow count how many bars back it was.

I hate At'ing people, but if you've got time and are willing, maybe @roar might have an idea?

Thanks to whomever has a solution!

I'm leaving this here in case someone else runs across it and needs help with something similar.

I decided to create a custom loop. Roar discussed them with me very in depth over the last few days on this thread if you wish to further your knowledge:

https://fxdreema.com/forum/topic/10894/scan-historical-data-for-conditions-values?page=1

Basically, my solution I decided on was similar:

Set up a variable "loopid" and set up a variable "candlelookbackID" and a constant "CandleLookback

After my other trade conditions (this is where I'd want to check if the price had violated the MA), I set up condition "loopid <= CandleLookback"

If true, it adds 1 to loopid and adds one to candlelookbackID (both are defaulted to 0).

Next block, condition, for longs, Candle Low > MA. In the candleID slot, I put in candlelookbackID for both of these.

If true, then the output goes back to the "loopid <= CandleLookback". This adds 1 more onto each of my variables and keeps running, one candle back until it hits a candle where the low is below the MA (breaks the loop) or until the loopid is greater than my CandleLookback.

Therefore, if loopid, when that first block is false (loopid <= CandleLookback), then it places my trade.

There's an additional block down there, after the buy/sell that resets loopid and candlelookbackID to 0 for the next time.

I hope this can be of help to someone else!

Thanks again for all your help, Roar (if you see this)

Bonus points if the method we conclude with has an optimizable input (so I can check if 3 candles is superior to 8 candles, for example).

I've been further experimenting with it (just ran a little EA with a node that has, for buys/above MA, Candle Low, Cross Below, Cross Width: 5, then my MA settings. I took the false output from that. Idea being that it only trades when price has NOT crossed below the MA within the last 5 candles).

I'm starting to feel like I need to create a custom loop that searches for the last candle that cross above/below, take the candle ID from that candle, and check that it is greater than 5? (Because if it's candleID 7, it's been more than 5 candles?)

Is this the most proper/correct/only way to accomplish this task? I simply need some logic that allows trades when price hasn't interacted with the MA in X candles (touch or cross).

Maybe this is less of a bug and more of a lacking feature, but:

Shouldn't we be able to enter a variable or a constant into the slot for "cross width" on the condition screens and such?

I have a project idea, but I need the logic "price has not touched/crossed an MA in the last X candles (could be 5 or whatever. Not particularly relevant at the moment).

If I use the cross above block, and change the cross width to 5, then take the false output from that, is that really all it takes?

So, for longs, Candle Low of CandleID 1 (starting on previous candle), cross above, width 5, take the false output.

I'm kinda having a hard time with the "cross width" idea mainly. It means that it's crossed above at any point in the last X candles, right?

Basically, if price is above, and the low touches and penetrates the MA, even if the open and close are above, I would like it to invalidate.

I'm sorry this question probably seems so simple, but the cross width functionality, though I think I clearly understand it, I'm not feeling confident in that understanding for whatever reason.

Thanks, peeps!

Thanks for the additional info, @roar

As you've probably noticed, MT4 is way more popular with brokers (practically every broker has it, whereas MT5 is a bit more scarce. You might be right. There could be some additional costs associated with supporting it at the broker level, or it could simply be more crap that can break and, if demand isn't high enough, then they don't bother with the added headache.

I decided to give MT5 a download and mannnnnn, there's soooooo much extra stuff int he strategy tester and stuff (plus I got a pretty beefy computer, and it allows me to use the processor to the best of its ability, which really speeds things up, also). I'm very pleased with it so far, to say the least.

I think a lot of people stick with MT4 for custom indicators since it's old an faithful (maybe for hedge positions, since I don't believe you can open conflicting positions on MT5? at least that's what I've heard)

Years back, I was interested in indicators primarily (I thought they were going to be the key to trading success), and I heard MT4 had way more and that they weren't compatible, so, after loading up both applications one time, I saw they graphically looked really similar, also, so I just assumed that MT4 was pretty much the same and I never looked back. I think at that point, I probably made the right decision.

Now that I'm building EAs and venturing down that route, I think the tables have turned. I'm using less custom indicators than ever and that added functionality of MT5 seems very worth it in my case. Thanks for the encouragement!

Thanks again for all your help, as well!

@l-andorrà Hahahaha. I'm kinda lurking on this thread hoping to hear what they say, also. Hopefully the OP might try to summon one of them. I'd love to get answers on this. I love the "on tester" stuff in MT4. I have a particular indicator backtesting EA that utilizes it and I can get some pretty good info out of it sometimes. I'd love to be able to incorporate this into my own EA builds.

If roar/miro come along, would one of you mind making a proper tutorial for it and labeling it as such and posting it in the tutorial forum, then linking it here so we know to check it? I'm sure it's worthy of that, if you have the time/energy/patience to do so.

@jsauter86 No worries, bud!

@roar

Sorry. Forgot to tag you again. Hahaha.

Gotcha! Yes! The Current is Bull was set to -1 and for some reason, while trying to take it all in, I thought it was the streak one in hindsight. I can completely see why it needs to be something other than 1 or 0 now that you've clarified it.

And that makes sense on the streak adjustment thing, too, now that you've pointed it out. The first time it detects the pattern (two simultaneous candles the same direction), the streak isn't 1, it's 2. Hahaha. So, that makes sense now that you've phrased it that way.

Ah! You make a good point that I didn't even think about--that it's counting from right to left, because as loopid changes, the candle number is increasing (and they increase from right to left, as you said) hahahaha. Now I've got to reorient my mind about how all this stuff works again, because I was instinctively working from left to right whenever I was trying to make sense of this stuff originally. Oof.

Thanks again for all the clarification! It's been really helpful! I don't think I have any other pressing questions at the moment. That might change as I dig further in.

There isn't any critical differences in how this works for MT5 coding or anything, is there?

I've not messed with MT5 in FxDreema at all, but I've noticed in some of your tutorials and stuff that you've used MT5 coding and I feel like I've seen Miro and some others using MT5 also. It's leading me to the assumption that you guys are comfortable in MT4 and helping others in MT4 since it's the most commonly used of the two (I think), but it seems like you guys probably prefer MT5 (otherwise you would present project ideas and stuff in MT4 only since it's the most common).

The reason I ask is because, from starting to look into MT5, it seems much more robust for EA building/testing, which is super exciting to me. I'm still early enough in this EA building journey to transition over this way, I guess I'm just curious if any major part of this doesn't apply to MT5 FxDreema building? Surely it's all pretty much the same on the FxDreema end of things (and different on the custom scripts and stuff).

Thanks for the additional info on arrays, too, btw. That's very beneficial. I've spent a lot of time in Excel previously, so I kind of see arrays as a table now, in essence.