MQL5 Help as One File:
mql5.chm mql5.pdf English

# Log-normal distribution

This section contains functions for working with log-normal distribution. They allow to calculate density, probability, quantiles and to generate pseudo-random numbers distributed according to the log-normal law. The log-normal distribution is defined by the following formula:

where:

• x – value of the random variable
• μ – logarithm of the expected value
• σ – logarithm of the root-mean-square deviation

In addition to the calculation of the individual random variables, the library also implements the ability to work with arrays of random variables.

 Function Description MathProbabilityDensityLognormal Calculates the probability density function of the log-normal distribution MathCumulativeDistributionLognormal Calculates the value of the log-normal probability distribution function MathQuantileLognormal Calculates the value of the inverse log-normal distribution function for the specified probability MathRandomLognormal Generates a pseudorandom variable/array of pseudorandom variables distributed according to the log-normal law MathMomentsLognormal Calculates the theoretical numerical values of the first 4 moments of the log-normal distribution

Example:

 #include  #include  #include  #property script_show_inputs //--- input parameters input double mean_value=1.0;  // logarithm of the expected value (log mean) input double std_dev=0.25;    // logarithm of the root-mean-square deviation (log standard deviation) //+------------------------------------------------------------------+ //| Script program start function                                    | //+------------------------------------------------------------------+ void OnStart()   { //--- hide the price chart    ChartSetInteger(0,CHART_SHOW,false); //--- initialize the random number generator      MathSrand(GetTickCount()); //--- generate a sample of the random variable    long chart=0;    string name="GraphicNormal";    int n=1000000;       // the number of values in the sample    int ncells=51;       // the number of intervals in the histogram    double x[];          // centers of the histogram intervals    double y[];          // the number of values from the sample falling within the interval    double data[];       // sample of random values    double max,min;      // the maximum and minimum values in the sample //--- obtain a sample from the log-normal distribution    MathRandomLognormal(mean_value,std_dev,n,data); //--- calculate the data to plot the histogram    CalculateHistogramArray(data,x,y,max,min,ncells); //--- obtain the sequence boundaries and the step for plotting the theoretical curve    double step;    GetMaxMinStepValues(max,min,step);    step=MathMin(step,(max-min)/ncells);    //--- obtain the theoretically calculated data at the interval of [min,max]    double x2[];    double y2[];    MathSequence(min,max,step,x2);    MathProbabilityDensityLognormal(x2,mean_value,std_dev,false,y2); //--- set the scale    double theor_max=y2[ArrayMaximum(y2)];    double sample_max=y[ArrayMaximum(y)];    double k=sample_max/theor_max;    for(int i=0; i=cells) ind=cells-1;       frequency[ind]++;      }    return (true);   } //+------------------------------------------------------------------+ //|  Calculates values for sequence generation                       | //+------------------------------------------------------------------+ void GetMaxMinStepValues(double &maxv,double &minv,double &stepv)   { //--- calculate the absolute range of the sequence to obtain the precision of normalization    double range=MathAbs(maxv-minv);    int degree=(int)MathRound(MathLog10(range)); //--- normalize the maximum and minimum values to the specified precision    maxv=NormalizeDouble(maxv,degree);    minv=NormalizeDouble(minv,degree); //--- sequence generation step is also set based on the specified precision    stepv=NormalizeDouble(MathPow(10,-degree),degree);    if((maxv-minv)/stepv<10)       stepv/=10.;   }

Updated: 2017.02.06