Why does my graph get sent to the end of the html when I publish it?

I'm using MATLAB R2021b and am trying to format my homework to output the graphs within the document, but when I try to plot something with a function inside it it throws all my graphs at the end of the publication rather than in line where I set it.

Here is my code

%% Module 1 MATLAB Exercise
%% Part 1
% Compute and plot the discrete distribution for a Binomially % distributed
% random variable n where the number of trials N=10 and p=0.7. The MATLAB
% function "nchoosek.m” is helpful.*

N = 10;     %Number of Trials

p = 0.7;    %Probability

q = 1-p;

k = (0:10); %Random variable from 0 to 10

bar(k,arrayfun(@(x) nchoosek(N, x)*p^x*q^(N-x), k));

snapnow;

%%
% The bar graph is centered around 7 successes and falls off on either
% side and though it is extremely small, there still is a very small
% possibility that there will be 0 successes.

%% Part 2
% In a for loop run M times, perform 10 Bernoulli trials sum to get
% the number of successes for each set of 10 trials. When done, use the
% MATLAB "hist.m" function to plot the tabulated sums with 11 bins
% specified to "hist.m” for binning the the numbers 0 to 10. You should
% normalize the histogram so that it sums to one, i.e., divide by M. Do
% this for M=10, 100, and 1000000, plot and compare.*

%%
% Histogram of 10 Bernoulli Trials

histogram(TenBernoulliTrials(10),11,'BinEdges',0:10,'Normalization','probability');

snapnow;

%%
% At just 10 trials, the total varies greatly, sometimes matching the shape
% of the graph generated in Part 1 pretty well, sometimes looking very
% different. This is because 10 is barely enough trials to look like the
% first graph. There are not enough data points for there to be anything
% that goes beyond barely resembling the graph from part 1. This is because
% there are not enough trials at this point for there to be a clear
% distinction between the different results.

%%
% Histogram of 100 Bernoulli Trials

histogram(TenBernoulliTrials(100),11,'BinEdges',0:10,'Normalization','probability');

snapnow;

%%
% At 100 trials, the shape starts to fit the graph from part 1 much better,
% though there are some minute differences that can be seen in certain
% sections of the histogram.The distribution can clearly be seen, but the
% actual numbers do not have the resolution of the bar graph from part 1.
% This is because there are enough trials at this point for there to be a
% clear distinction between the different results, but not enough to
% accurately define the probability.

%%
% Histogram of 1000000 Bernoulli Trials

histogram(TenBernoulliTrials(1000000),11,'BinEdges',0:10,'Normalization','probability');

snapnow;

%%
% At 1000000 trials, the shape fits nearly exactly with the bar graph
% generated from Part 1. The shape of the graph matches nearly perfectly
% and the magnitudes of the normalized bar graph match the probability
% distribution from Part 1. This is because there are enough trials at this
% point to have a small enough resolution to generate values that are
% comparable to the first part.

%% Bernoulli Trial Function
function SUM = TenBernoulliTrials(M)

    SUM = zeros(M,1);

    for x = 1:M
        for N = 1:10
        
            if rand < 0.7
        
                SUM(x) = SUM(x) + 1;
        
            end
        end
        
    end

end

%%
% This is the function used to generate the histographs. It uses a random
% number generator to simulate the 0.7 probability for a success and sums
% the number of successes M number of times which is the output of the
% function.

And here's the ending of what I get when I publish it

I've attached the pdf of the full document

Answer 1

[Ive J]

I cannot reproduce this behavior.

[Jason Yoon]

The pdf you attached replicates the behavior exactly, the two graphs are at the end rather then when they are called