stat645
Project 1.
The idea of project 1 is to give you some practice applying what you've learned by creating a new visualisation. You have two options: extend an existing visualisation for a new situation, or build a visualisation for a particular type of model
Extend an existing visualisation:
- extend the hanging rootogram to 2d
- extend the hanging rootogram to work with continuous density estimates
- create a general method for converting 2d joint density plots into
conditional density plots
- any of your own ideas, after discussion with me
Create a visualisation method for a specific type of model:
- a linear model with two continuous and one categorical predictors
- hierarchical clustering
- model based clustering from mclust
- survival analysis models
- any of your own ideas, after discussion with me
For these projects you'd need to implement two functions: one that turns the model object into a data frame with a useful set of columns, one that takes that data frame and produces a default visualisation, with a small selection of useful parameters.
Outputs
- by the end of the week: an idea, know who you want to work with, and email me to set up a meeting
- feb 21: a 5 page report (should be mostly graphics) that describes what you did, and what your reasoning was, along with the code that you used.