Improve argument handling in model functions

skent259 / mildsvm

Multiple Instance Learning with Distributions, SVM

Other

3 stars 0 forks source link

Improve argument handling in model functions #34

Closed skent259 closed 2 years ago

skent259 commented 3 years ago

There is a lot of repeated code that is copied and pasted to start most functions. As the complexity of this package grows, I think it might be helpful to have a standard argument handing function, with some potential sub-functions.

This may not shorten the code (such checking is likely annoying to do in R), but it could make things more manageable in the future.

skent259 commented 2 years ago

This could massively reduce the testing burden of the package, because of the reduced conditional logic required in each function. Separate tests for argument handling would be ideal.

skent259 commented 2 years ago

Some improvements in #44

skent259 commented 2 years ago

Some big items left to do

refactor mismm.default()
refactor smm.default()
refactor predict.smm()
refactor _fit functions, including selection of instances, selection max warning

skent259 commented 2 years ago

Major improvements in #58, closing for now