Population age model "randi" distribution

Does anyone know how I can randomly generate numbers that will be distributed to a population-by-age distribution? I want to be able to input a population number, say 1000, and have it return 1000 random numbers distributed in a curve that looks something like the USA age distribution curve... Please explain all code

Thanks!

[Torsten]

And where is the US age distribution curve ?

[Pesach Nestlebaum]

Something along these lines:

https://knoema.com/infographics/egyydzc/us-population-by-age-and-generation-in-2020

Alternatively, maybe I can assign portions of the randomly generated numbers to specific percentages? For example, I want to assign numbers 40-50 to make up 45% of the generated numbers

Answer 1

[Umang Pandey]

Hi Pesach,

As per my understanding it appears that you are trying to generate random numbers from any desired distribution, in your case "The US Age distribution curve".

Here’s an example that can be followed:

% Assuming some Example Data 
age_groups = 0:100; % Age groups 
population = rand(1, 101); % Dummy population data 
% Normalize the data to get the PDF 
pdf = population / sum(population); 
% Calculate the CDF 
cdf = cumsum(pdf); 
% Generate uniform random numbers 
N = 1000; % Number of random numbers to generate 
u = rand(N, 1); 
% Sort the CDF and age_groups 
[cdf, idx] = sort(cdf); 
age_groups = age_groups(idx); 
% Find the corresponding ages using the CDF 
random_ages = interp1(cdf, age_groups, u, 'nearest'); 

To know more about the “interp1” function, you can refer the following documentation:

https://www.mathworks.com/help/matlab/ref/interp1.html

Best,

Umang