Histogram to a CDF/PDF

Hello, This is a screenshot of a table I have constructed for work.

Just to play it safe, I blacked out the column names, though it would be hard to assume anything with just 7 rows of the table to go off of. We will call the 5 fields "column1, column2, etc."

So I am able to create the hisogram of any of the columns, besides 3, but that isn't needed because it is all '94'.

I do:

histogram([a(1:135756).column1])

and the histogram works perfectly.

How would I do a CDF or PDF of this data?

I have tried:

histogram([a(1:135756).column1],'Normalization',pdf)

or

histogram([a(1:135756).column1],'Normalization',cdf)

but nothing changes from the original histogram.

Thank you!

[Adam Danz]

histfit(data,nbins,dist)

Answer 1

[Bruno Luong]

A=[a(1:135756).column1];

figure
subplot(2,1,1);
histogram(A,'Normalization','pdf');
ylabel('pdf');
subplot(2,1,2);
histogram(A,'Normalization','cdf');
ylabel('cdf');

[Sclay748]

ahhhh beautiful.

Thank you!

Answer 2

[Alan Stevens]

You can get a CDF as follows:

% Modified Kaplan-Meier CDF
% assumes each point is representative of 1/N of the population.

a = sort(a(:,1)); % so all the data for a are sorted in ascending order
N = length(a);
for k = 1:N
    CDF(k) = (k - 0.5)/N;
end
plot(a,CDF)

Because you have a large number of points you could simply numerically differentiate the CDF to get a PDF.

[Sclay748]

hi Alan, I got an error for using sort. Says incorrect argument data type in call to function 'sort'.

[Alan Stevens]

I've asumed a is just a column vector. If it's a different structure of some sort you will need either to convert it to a simple vector, or find some way of sorting the more complicated structure.

[Bruno Luong]

Hmm it cries for replacing the for-loop

a1 = sort([a(1:135756).column1]);
N = length(a1);
CDF = (0.5:N-0.5) / N;
plot(a1, CDF);

[Sclay748]

Bruno, that worked. Is it supposed to be a single curved line?

I thought it would still plot the bars, but arranged by CDF. Or that is atleast how my boss's turned out when he showed me an example using histogram(......,'Normalization',cdf)