JIYIK CN >

We will study different ways to plot frequency distribution curve of data in MATLAB. We will use different sample codes and relevant outputs to clear your concepts and provide you complete insights using MATLAB.

Note that MATLAB allows users to perform statistical operations using different functions such as histfit(), , hist()and sshist().

A frequency distribution is a list/table/graph that shows how often certain possibilities occur in our data set. We can use MATLAB to plot a frequency distribution curve for our data.

We can add data manually or import it from an Excel worksheet. Using different functions, we can easily get the desired curve.

We use the function in MATLAB sshist()to calculate the density estimate and obtain the frequency curve of our data.

To plot a frequency distribution graph in MATLAB, we first need to calculate the optimal number of bins for the histogram.

For our 核密度estimation calculations, we will assume that our frequency distribution has the standard bin length for histograms in MATLAB.

However, this approach is probably unrealistic. The first approach calculates the optimal bin width to produce the sharpest curve while preserving as much of the core PDF as possible.

The function sshistis not included in the list of standard functions of MATLAB, but we can obtain the source code of the function and use it to run our code to calculate the density estimate of bins. See the example,

Our_data = randn(1,1e4);

optN = sshist(Our_data);

figure(1)
[M,C] = hist(Our_data);
[Nop,Cop] = hist(Our_data,optN);
[f,xi] = ksdensity(Our_data,Cop);

dNom = mode(diff(Center));
dNom = mode(diff(Cop));

plot(Center,N/dNom,'.-',Cop,Nop/dNom,'.-',xi,f*length(Our_data),'.-')
legend('Default','Optimum','ksdensity')
title('Frequency Distribution')

Output:

As mentioned above, we first calculate optimal-bin. The reason behind this calculation is to provide structure to our frequency distribution histogram.

When we use ksdensitythe function, it returns a flat curve. So, it depends on the nature of our work and the type of output graph we want based on our data.

Note that we used an additional m-file to successfully execute this code. The function sshist()is not predefined in MATLAB.

That's why we need to copy the source code of this function, create an m-file and paste the copied code here.

Then execute the code to get the frequency distribution graph.

In MATLAB we use histfit()the function to obtain the frequency curve of our data.

histfit()To construct a frequency distribution histogram with a normal distribution fit using , let us take the data sample size to be 100, the variance to be 1, and the mean to be 10.

rng default;
freq_dist_data = normrnd(10,1,120,1);
histfit(freq_dist_data)

Output:

Note that histfitthe function in MATLAB fitdistworks similarly to the function to fit our frequency distribution to our dataset and return all the values ​​used in the fit.

fitting_values = fitdist(freq_dist_data,'Normal')

Output:

fitting_values = 

  NormalDistribution

  Normal distribution
       mu = 10.0894   [9.88181, 10.297]
    sigma =  1.1484   [1.01919, 1.31541]