This tool is designed to provide a smoother, more refined view of market price action by filtering out the "noise" typically found in standard financial charts. Instead of simply showing raw price jumps, it uses a mathematical approach—known as a Kalman filter logic—to estimate the underlying trend of an asset.
Core Functionality
The script works by analyzing historical price data to predict the next logical step in the market's movement. It doesn't just react to what has happened; it attempts to calculate the "velocity" and "distance" of price changes to create a fluid line that represents the true direction of the market.
Key Components
- Data Selection: The tool allows users to choose which type of market data it should analyze. You can set it to follow the closing price, opening price, highs, lows, or various weighted averages (such as a "Typical" price which balances high, low, and close). By default, it is set to use a weighted calculation that gives significant importance to the current closing price.
- Sensitivity Controls: There are specific parameters that allow you to adjust how "sharp" or responsive the filter is. This helps in balancing the trade-off between a smooth line that lags slightly behind the price and a reactive line that might be too jittery.
- Dynamic Trend Visualization: The tool produces two distinct visual paths on your chart:
- When the calculated momentum is positive (moving upward), it displays an upward trend line.
- When the momentum shifts to negative (moving downward), it switches to a downward trend line.
How it Works in Practice
The logic processes data from the past toward the present. At each step, it calculates how far the current price has drifted from its previous expected position. It then uses this "error" and the current "velocity" to refine its next prediction. By continuously updating these variables, the indicator produces a continuous flow that highlights the primary trend while smoothing out minor, insignificant fluctuations in the market.
//+------------------------------------------------------------------+
//| Kalman filter.mq4 |
//| Copyright © 2006, iziogas@mail.com. |
//+------------------------------------------------------------------+
#property copyright "iziogas@mail.com"
//----
#property indicator_chart_window
#property indicator_buffers 2
#property indicator_color1 Turquoise
#property indicator_color2 Orange
//---- indicator parameters
//Mode description
// 0: Close
// 1: Open
// 2: High
// 3: Low
// 4: Median (H+L/2)
// 5: Typical (H+L+C/3)
// 6: Weighted (H+L+C+C/4)
extern int Mode=6;
extern double K=1;
extern double Sharpness=1;
extern int draw_begin=500;
//---- indicator buffers
double ExtMapBufferUp[];
double ExtMapBufferDown[];
//----
int ExtCountedBars=0;
//+------------------------------------------------------------------+
//| Custom indicator initialization function |
//+------------------------------------------------------------------+
int init()
{
//---- drawing settings
SetIndexStyle(0,DRAW_LINE);
SetIndexStyle(1,DRAW_LINE);
IndicatorDigits(MarketInfo(Symbol(),MODE_DIGITS));
SetIndexDrawBegin(0,draw_begin);
SetIndexDrawBegin(1,draw_begin);
//---- indicator buffers mapping
SetIndexBuffer(0,ExtMapBufferUp);
SetIndexBuffer(1,ExtMapBufferDown);
//---- initialization done
return(0);
}
//+------------------------------------------------------------------+
//| |
//+------------------------------------------------------------------+
double iValue(int mode, int shift)
{
switch(mode)
{
case 0:
return(iClose(NULL,0,shift));
case 1:
return(iOpen(NULL,0,shift));
case 2:
return(iHigh(NULL,0,shift));
case 3:
return(iLow(NULL,0,shift));
case 4:
return((iHigh(NULL,0,shift)+iLow(NULL,0,shift))/2);
case 5:
return((iHigh(NULL,0,shift)+iLow(NULL,0,shift)+iClose(NULL,0,shift))/3);
case 6:
return((iHigh(NULL,0,shift)+iLow(NULL,0,shift)+iClose(NULL,0,shift)+iClose(NULL,0,shift))/4);
default:
return(0);
}
}
//+------------------------------------------------------------------+
//| |
//+------------------------------------------------------------------+
int start()
{
if(Bars<=draw_begin) return(0);
//----
int i;
//----
double Velocity=0;
double Distance=0;
double Error=0;
double value=iValue(Mode,draw_begin+1);
//----
for(i=draw_begin;i>=0;i--)
{
Distance=iValue(Mode,i) - value;
Error=value + Distance * MathSqrt(Sharpness*K/100);
Velocity=Velocity + Distance*K/100;
value=Error+Velocity;
//color lines
if (Velocity>0)
{
ExtMapBufferUp[i]=value;
//ExtMapBufferUp[i] = S;
ExtMapBufferDown[i]=EMPTY_VALUE;
//----
if (ExtMapBufferUp[i+1]==EMPTY_VALUE) ExtMapBufferUp[i+1]=ExtMapBufferDown[i+1];
}
else
{
ExtMapBufferUp[i]=EMPTY_VALUE;
ExtMapBufferDown[i]=value;
//ExtMapBufferDown[i] = S;
if (ExtMapBufferDown[i+1]==EMPTY_VALUE) ExtMapBufferDown[i+1]=ExtMapBufferUp[i+1];
}
}
//---- done
return(0);
}
//+------------------------------------------------------------------+
Comments