Understanding Probability Distributions for Generating Random Data

[EliteLegend]

I have recently come across the 80/20 rule... I am using the Pareto Distribution's pdf to generate some dataset that I wanted... Now if I have a set of 50 items and I need to generate 40% of these items 60% of the times, how am I supposed to go about doing this? I know how to select items with certain probabilities but this task is confusing me... Anyone has some inputs for me please?

 

[EnumaElish]

One way is to create duplicate (multiplicate) items until you reach the desired proportion, then use a uniform rule to select.

For example, if I had 2 items {x, y} and wanted to obtain x 67% of the time, I'd duplicate x once, and make draws from the set {x, x, y}.

 

[mXSCNT]

To select one of m items, from a total of n items, a proportion y of the time, you need to select each of the m items with probability y/m, and each of the remaining n-m items with probability (1-y)/(n-m)