Week 5 practice

Prediction modelling

In Week 5 Workshop we are introducing classification and regression models for prediction.

What is random?

In the week 4 interactable code we introduced the idea of random sampling when we generated a dataset of clusters.

Setting a seed in random number generation will ensure that the sequence of numbers generated by the random number generator is reproducible.

Here are examples:

Randomly sampling a number between 0 and 1

runif will randomly select from a uniform distribution (random uniform). runif(1) will generate a random number with equal probability from a uniform distribution between 0 and 1. Try changing the seeds

Randomly sampling in sequence

The seed initialises the random number generator to a known state. Each time a seed is set to the same value, the sequence of random numbers generated after that point will be the same.

Try changing the seeds:

Random and regular fields of points

The random field of points from Lecture 5. d1 is random points from the random uniform distribution:

To create the evenly distributed bit (d2), you have to start with an even grid of points:

Then add a little bit of random noise to it:

Paste them together and plot it:

Coin flips

We haven’t set a seed here - each time you run this cell, you will get different results. Try setting a seed.

Roll a dice

We’ve set our seed to the current time! Why might we do this?

The default random number generator for R is the Mersenne-Twister. https://en.wikipedia.org/wiki/Mersenne_Twister