@vv  Thank you very much!

@Markiyan Hirnyk Thank you very much!

@vv My worksheet Distribution2.mw
I would like use frequency absolute values, not relative.

However, if I use the necessary parameters for me (bincount = 10, frequencyscale = absolute) gamma distribution loses its shape. How I can avoided this problem?
 

h := Histogram(a, bincount = 50, frequencyscale = absolute, range = 0 .. 5);
with(Statistics);
m := Mean(a); v := Variance(a);
b := v/m; c := m/b;
P := DensityPlot(RandomVariable(Gamma(b, c)));
plots[display](h, P);

@vv

@Markiyan Hirnyk My second question about using gamma distribution disappeared. Perhaps it was corrected or deleted due to the large size with a lot of numbers as an example?

@vv Thank you very much!
1. Now I can determine the analytic function obtained distribution like that?

Can I find now for example Variance(P), and Kurtosis(P)?

@Christopher2222  can I determine the analytic function obtained distribution like that ?

?

@Markiyan Hirnyk 

with(Student[Statistics]);
Ob1 := [30, 30.5, 31, 31.5, 32, 32.5, 32.6, 33, 33.1, 33.3, 33.6, 34, 35, 36];
Ex1 := [20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20];
ChiSquareGoodnessOfFitTest(Ob1, Ex1, level = 0.5e-1);

maybe so?

@Markiyan Hirnyk When i asked about smoothing the histogram I mean method  that offered by Christopher2222 4301. But I did not consider chi-square test..  
How can I apply this method ( chi-square test ) to determine the most appropriate type of distribution?

@Christopher2222 

1) Yes, It is very suitable, thank you! 

2) So, first of all,  I need check chi-square test?

3) And how can I determine the analytic function obtained distribution?