How to create a uniform distribution in R?
This recipe helps you create a uniform distribution in R
Recipe Objective
Random numbers are generated in quite a few cases in statistics to carry out sampling and simulation. Mostly, a data scientist is in a need of a set of random numbers which are mostly taken from two types of distribution:
- Uniform distribution
- Normal distribultion
These random numbers generated mimic the properties of uniform or normal distribution in a certain interval.
In this recipe, you will learn how to create a random uniform distribution..
Uniform distribution is a type of probability distribution in which all the numeric variables have an equal probability to occur. The are the most popular type of distribution in generating random numbers.
Note: Whenever we are generating random numbers, you are using an algorithm that requires a seed whose function is to initialise. These numbers are actually pseudorandom numbers which can be predicted if we know the seed and the generator. Setting a seed means iniltialising a pseudorandom generator. We set a seed when we need the same output of numbers everytime you want to generate random numbers. If we don't set a seed, the generated pseudorandom numbers are different on each execution.
Example: Creating a uniform distribution by generating 100 random numbers from a uniform distribution by seeting a seed
We use runif() function to carry out this task.
Syntax: runif(n, min = , max = )
where:
- n = size of the distribution
- min, max = specifies the interval in which you would like the distribution to be
Additionally, use set.seed() function to set a seed. We specify any integer in the function as a seed.
# setting a seed
set.seed(20)
# using random numbers from normal distribution between 1 and 30
uniform_dist = runif(100, min = 1, max = 30)
#plotting a histogram of the generated numbers using hist() function
hist(uniform_dist, breaks = 3)
Note:
- The distribution remains constant even after multiple execution.
- You can see that the mean, mode and median co-incides in the above plot indicating a normal distribution
What Users are saying..
Ray han
I think that they are fantastic. I attended Yale and Stanford and have worked at Honeywell,Oracle, and Arthur Andersen(Accenture) in the US. I have taken Big Data and Hadoop,NoSQL, Spark, Hadoop... Read More