A trading system without a drain. Needs a programmer. - page 3

New comment
 
EVladMih писал (а):

...register the simultaneous presence of SEVERAL stochastics on several TFs in the correct positions.

... I.e. the system only gives an entry, but gives it error-free.

This task is trivial for any MQL-programmer, if the author can formulate the idea in figures, not only for 3 timeframes, but for at least 7, from 1 minute to two-day periods. I dare say that when the idea is implemented (no matter by whom), that is when we will see (when testing on the history) that, along with correct entries, the tester will find a lot of situations that satisfy conditions but lead to exactly the opposite results. And it is good, if there are significantly more positive inputs.
That's when optimization and fitting the parameters to the history with all its consequences will start ... This is one of the advantages of testing on the history: it helps to destroy unrealizable illusions that look so beautiful when looking at charts and are proved by one or two months of manual work on the demo.
Although, as an auxiliary tool for manual work, it could well be applied.
 
YraZ, I've had the opposite result. Entries are frequent and bad. Can you elaborate a little more ?
 
Valmars писал (а):
EVladMih wrote (a):

...register the simultaneous presence of SEVERAL stochastics on several TFs in the correct positions.

... I.e. the system only gives an entry, but gives it error-free.

In general, the problem is trivial for any MQL-programmer, if the author can formulate the idea in figures, not only for 3 timeframes, but for seven, from minutes to days. I dare say that when the idea is implemented (no matter by whom), that is when we will see (when testing on the history) that, along with correct entries, the tester will find a lot of situations that satisfy conditions but lead to exactly the opposite results. And it is good, if there are significantly more positive inputs.
That's when optimization and fitting the parameters to the history with all its consequences will start ... This is one of the advantages of testing on the history: it helps to destroy unrealizable illusions that look so beautiful when looking at charts and are proved by one or two months of manual work on the demo.
Although, as an auxiliary tool for manual work, it could well be applied.

Amateurs often move progress precisely because they do not know that according to some clever laws it is impossible.

There is no desire (and no time) to engage in argument, I want to pay attention - how many people appear at any work and lively start to disassemble it for parts, pick it up with a screwdriver, chop with an axe, shit all over it, explain that it is impossible, because it can never be, etc. etc.

In doing so, these well-wishers don't even notice that they are often objecting to themselves....
After all, the ad says NOTHING about the experts!!!!
It's just an indicator that signals the coincidence of the required parameters on different TFs! That's all!!!

2 Valmars: Don't take it personally, this is a generalized post - I often see that everywhere. I`m not the only one who gets bombarded, but even very famous traders. The answer to you personally is the first line (about dilettantes, i.e. me) :)
 
eugenk1 писал (а):
YraZ, I've had the opposite result. Entries are frequent and bad. Can you elaborate a little more ?

here's a report of my inputs
on my signal
the inputs are GREAT ! the whole secret is in the stoch parameters and the combination of signals

the problem is with the outputs! here i just put a short TP to show that there are no left handed inputs


that's why when he says it works I have no doubts

because here's an example.
but the problem is I haven't found a combination
that can be entered more often and I don't have an exit criterion either.

In general, stochastic is great in a flat!

The author apparently has a TS on STOCK
Files:
steit.zip  11 kb
 
YuraZ писал (а):
eugenk1 wrote (a):
YraZ, I've had the opposite result. Inputs are frequent and bad. Can you elaborate a bit more ?

Here's a report of my inputs in general the stochastic is awesome in the flat!
the author seems to have a stochastic TS

It's nice to deal with a person "in the know". :) I hope I have a letter from YuraZ in my mailbox.
The flat flat is great, and when the trend is visible with the naked eye it is a "filler".
It can give exit only in flat (even flips), and in general for the stops we need other brakes.

The guys on this forum are not simple and don't want to take the easy way out!
I have been informed today that the emails from some of them have gone to a completely different country and have been mistaken for spam.
The reason: Instead of my current address (they change sometimes, sometimes they die), which I have twice posted openly, dug out apparently from the profile of the old address.
I'm giving again the current one (send it to anyone who needs it): f-x{}fxmail.ru

I'll tell you about Stochastic, but some other time. I'm in a bit of a jam at the moment.
 
It may come in handy (?), I have a simple globalist for getting data from several TFs - I already wrote here 'How to combine two indicators?
In the indicators you give the global variables of an indicator, or its "power" (2,1,0,-1,-2) - run the globalist on any TF and get the joint charts.
In your case (if I understand correctly) it will be like this:
//---- В ИНДИКАТОРЕ "СТОХАСТИК"
...
...
...
string        ThisName;
//---------------------------------------------------------------------
void deinit()
{
   if(GlobalVariableCheck(ThisName)==true)
      GlobalVariableDel(ThisName);
   Comment("");
   return;
}
//---------------------------------------------------------------------
int init()
{
...
...
...
   ThisName=Symbol()+"_M"+Period()+"_STOH";
   return(0);
}
//---------------------------------------------------------------------
int start()
{
...
...
...
   if (Pos==0) 
   {
      ST=0.0;
      if(BUF[0]>...) ST=1.0;
      if(BUF[0]<...) ST=-1.0;
      GlobalVariableSet(ThisName,ST);
   }
   return(...);
}
//---------------------------------------------------------------------
 
 
 
//---- В ИНДИКАТОРЕ "ГЛОБАЛИСТ"
//---------------------------------------------------------------------
//---------------------------------------------------------------------
int start()
{
double m5,m15,m30,m60,m240;
      if(GlobalVariableCheck(Symbol()+"_M5_STOH")==true)
         m5=GlobalVariableGet(Symbol()+"_M5_STOH");
      if(GlobalVariableCheck(Symbol()+"_M15_STOH")==true)
         m15=GlobalVariableGet(Symbol()+"_M15_STOH");
      if(GlobalVariableCheck(Symbol()+"_M30_STOH")==true)
         m30=GlobalVariableGet(Symbol()+"_M30_STOH");
      if(GlobalVariableCheck(Symbol()+"_M60_STOH")==true)
         m60=GlobalVariableGet(Symbol()+"_M60_STOH");
      if(GlobalVariableCheck(Symbol()+"_M240_STOH")==true)
         m240=GlobalVariableGet(Symbol()+"_M240_STOH");
      if(m5>0.5) m5=m5+0.05;
      if(m5<-0.5) m5=m5-0.05;
      if(m15>0.5) m15=m15+0.1;
      if(m15<-0.5) m15=m15-0.1;
      if(m30>0.5) m30=m30+0.15;
      if(m30<-0.5) m30=m30-0.15;
      if(m60>0.5) m60=m60+0.2;
      if(m60<-0.5) m60=m60-0.2;
      if(m240>0.5) m240=m240+0.25;
      if(m240<-0.5) m240=m240-0.25;
      Buf_M5[0]=m5;
      Buf_M15[0]=m15;
      Buf_M30[0]=m30;
      Buf_H1[0]=m60;
      Buf_H4[0]=m240;
}


The chart will have lines from all TFs with values 1 (sell), 0 (rest) or -1 (buy) plus/minus a bit, so they do not cover each other. Thus it is possible to put on one chart any indices (same/different) from any number of TF (but <=8 :) and to run the globalist on any TF, even on M1.

 
Bookkeeper писал (а):
It may come in handy (?), I have a simple globalist for getting data from several TFs - already wrote here 'How to combine two indicators?
... In your case (if I understood correctly) it would be like this:
...


The chart will have lines from all TFs with values 1 (sell), 0 (rest) or -1 (buy) plus/minus a bit, so they do not cover each other. Thus it is possible to put on one chart any indices (same/different) from any number of TF (but <=8 :) and to run the globalist on any TF, even on M1.


Unfortunately, it's all "cold" for me...
From the higher TFs will the lower ones be displayed correctly? I've heard that it is possible only vice versa.
And is it possible to correctly display this variant on the visualizer?
It is important to me, because we are working with small TFs and it takes a very long time to select parameters in real time.
 
EVladMih писал (а):
Bookkeeper wrote (a):
Might come in handy (?), I have a simple globalist for getting data from multiple TFs - already wrote here 'How to combine two indicators?'
... In your case (if I understand correctly) it would be like this:
...


The chart will have lines from all TFs with values 1 (sell), 0 (rest) or -1 (buy) plus/minus a bit, so they do not cover each other. Thus it is possible to put on one chart any indices (same/different) from any number of TF (but <=8 :) and to run the globalist on any TF, even on M1.


Unfortunately, this is all "cold" for me...
From the higher TFs will the lower ones be displayed correctly? I've heard it's only possible the other way round.
Only in this case it is YES. The only problem is that the Globalist has no history, only the period when all the indices are "working", i.e. while the terminal is on. About the "visuals" - I have no idea, ask the wise, I don't use EAs or tester.
What's cold for you?
If it's not clear how to include in an indicator - send any indicator to my email (in my profile) or post it here. It doesn't have to be yours, I'm not going to ask for your secrets :), I have enough crazy ideas of my own. I just want it to be similar to yours and then you will be able to transfer necessary code snippets to your place. And specify the condition that you need to buy/sell/seat-silently, like in RSI: >70 (go down) or <30 (go up) or 30...70 (wait) - in this very "simplified" version, so that you can then specify how you want it.
 
Sorry, turns out there's no username in my profile. Get it from CodeBase in the header of any of my scripts.
 
Valmars писал (а):
EVladMih wrote (a):

...register the simultaneous presence of SEVERAL stochastics on several TFs in the correct positions.

... I.e. the system only gives an entry, but gives it error-free.

In general, the problem is trivial for any MQL-programmer, if the author can formulate the idea in figures, not only for 3 timeframes, but for seven, from minutes to days. I dare say that when the idea is implemented (no matter by whom), that is when we will see (when testing on the history) that, along with correct entries, the tester will find a lot of situations that satisfy conditions but lead to exactly the opposite results. And it is good, if there are significantly more positive inputs.
That's when optimization and fitting the parameters to the history with all its consequences will start ... This is one of the advantages of testing on the history: it helps to destroy unrealizable illusions that look so beautiful when looking at charts confirmed by one or two months of manual work on the demo.
Although, as an auxiliary tool for manual work, it may well be applicable.

I'm not sure it is. Anyway, when I started using TFs I "bore" only one loss and that was a foolish move when I set SL=20p and after the trigger the price went where I ordered it. But results are really not so good... Too few deals, small profits... and generally boring :). I have nothing to boast about, but it makes me stronger (patience).
1234567
New comment