Trait rand_split::TTVSplit

pub trait TTVSplit<T>: Iterator<Item = T> + Sized {
    pub fn split_ttv(self, splits: [f32; 3]) -> TTVIterator<T, Self>;
}

An iterator type to generate train/test/validaton splits from a data stream of unknown lenght.

It is the same as PartsSplit but with a fixed size of 3 (it uses arrays instead of vectors).

See the examples on the repository for both iterator traits.

Required methods

pub fn split_ttv(self, splits: [f32; 3]) -> TTVIterator<T, Self>

splits contains weights that are not required to sum up to 1.

Loading content...

Implementors

impl<'a, T, I> TTVSplit<T> for I where
    I: Iterator<Item = T>, 

pub fn split_ttv(self, splits: [f32; 3]) -> TTVIterator<T, Self>

Loading content...