Put in a good word about the occasional wanderer... - page 23

New comment
 
The point is that the maximum and minimum can be searched independently, and so the input-output will be independent of each other.
 
Sorento:

here or elsewhere. because we don't generally know the number of "princes"/bars.

So the distribution of "total choice time" is what's interesting...

There's a lot of scriptwriters around here - maybe someone will make one.

;)

You're welcome! This is on EURUSD minutes, 250 thousand bars analysed.

Descriptive Statistics (BB_distr) Include condition: bb_length<300

 Valid N Mean Median Mode Frequency - of Mode Lower - Quartile Upper - Quartile Percentile - 10,00000 Percentile - 90,00000 Std.Dev.
bb_length 224214 33,75288 12,00000 1,000000 15464 5,000000 37,00000 2,000000 96,00000 52,27472


and here's a picture:



//+------------------------------------------------------------------+
//| script program start function                                    |
//+------------------------------------------------------------------+
int start()
  {
//----
   int i,j;
   int n;
   int h=FileOpen("BB_distr",FILE_WRITE|FILE_CSV);
   for(i=0;i<Bars;i++)
   {
      n=0;
      for(j=1;;j++)
      {
         n=j;
         if(i+j>=Bars) break;
         if(High[i+j+1]>=Close[i]&&High[i+j]<Close[i]||Low[i+j+1]<=Close[i]&&Low[i+j]>Close[i]) break;
      }
      FileWrite(h,n);
      FileFlush(h);
      if(i%1000==0) Print("Обработано баров: ",i," из ",Bars);
   }
   FileClose(h);
//----
   return(0);
  }
//+------------------------------------------------------------------+
 
alsu:

You're welcome! This is on the EURUSD minutes.

Descriptive Statistics (BB_distr) Include condition: bb_length<300

 Valid N Mean Median Mode Frequency - of Mode Lower - Quartile Upper - Quartile Percentile - 10,00000 Percentile - 90,00000 Std.Dev.
bb_length 224214 33,75288 12,00000 1,000000 15464 5,000000 37,00000 2,000000 96,00000 52,27472


and here's a picture:

Just a Miracle! As you'd expect...

Thank you!

The neophytes will be delighted

 
alexeymosc:

And where does the exponential distribution come from? How do you calculate the sum of the numbers?

Exponential, because the logarithm of the density of the distribution looks like a straight line with R^2=0.9997. So the assumption is quite reasonable, in my opinion.

In addition, I once gave (not strictly, though, but still) a justification for it to be exponential: in short, it is logically equivalent to a set of two assumptions: 1) the price movement is a reaction to external information and 2) the market is a statistical fractal, i.e. there is a distribution independent of scale.

 
Hmm. Everyone seems to have gone off to solve the princess and forex problem... I guess I'll do it then).
 

You got it all wrong.

Forex is a princess. When she thinks she's already "THERE HE IS !", she makes her choice and turns back around. :)

Another thing is that she is not very familiar with the task and especially with the right decision. Right.

So, the trader's tasks are:

1. To determine the competence and the current degree of maturity of the Princesa.

2. To see the queue of Princes a couple of members ahead.

3. To put a limiter in the pocket of a cute enough prince from the queue at the moment when the Princesa is ready for everything.

4. Wait for the U-turn while whistling Mendelssohn.

5. After the turn - see point 1.

;)

 
MetaDriver:

You got it all wrong.

Forex is a princess. When she thinks she's already "THERE HE IS !", she makes her choice and turns back around. :)

Another thing is that she is not very familiar with the task and especially with the right decision. Right.

So, the trader's tasks are:

1. To determine the competence and the current degree of maturity of the Princesa.

2. To see the queue of Princes a couple of members ahead.

3. To put a limiter in the pocket of a cute enough prince from the queue at the moment when the Princesa is ready for everything.

4. Wait for the U-turn while whistling Mendelssohn.

5. After the turn - see point 1.

;)

That is why I suggested to apply the strategy to MACD with 33. In doing so I suggest to consider that a move away from zero (as an option for one of the inputs), or better yet, a move away from extremum 33 to the level of two spreads - is already a fact of starting a new look. The task is complicated - the exact length of the resulting queue is not known in advance and therefore the drift of the average may be very considerable.

Don't forget about the arcsinus...

;)

 

princes on the graph where? prices, increments?

what is the prince's target function? For example, two princes - two prices. The higher the prince, the better? :)

 

The prince is the accumulated paper profit from the moment the prince has taken a pose (opened). The higher the profit, the better.

What else could it be? But princes are dependent, aren't they?

 
Avals:

princes on the chart where? prices, increments?

What is the princess' target function? For example, two princes - two prices. The higher the prince, the better? :)

prince is an extremum.

The princess reducing her prince requirement from the best to one of the top two has increased her chances from 36.8 to 57.4%.

Mathemat:

The prince is the accumulated paper profit from the moment the princesso took a pose (opened). The higher the profit, the better.

What else could it be? But the princes are dependent...

What makes you think that? A prince is any extremum on a bounded set.

and I'm tired of hearing the dependency mantra - then the market is bad because it's random and unsteady.

Now it's bad in that it's persistent...

;)

1...1617181920212223242526
New comment