Struct Dataset 

pub struct Dataset { /* private fields */ }

A labelled dataset stored column by column.

Before a search, every column must be sorted and its values indexed; see Dataset::is_prepared. Dataset::from_rows and the file reader do both.

Implementations

impl Dataset

pub fn new() -> Self

An empty dataset.

pub fn is_prepared(&self) -> bool

Whether both preparation steps have run. See SearchError::UnpreparedDataset.

pub fn count(&self) -> usize

Number of instances. Zero for an empty dataset.

pub fn is_empty(&self) -> bool

Whether the dataset has no instances.

pub fn num_features(&self) -> usize

Number of feature columns.

pub fn num_labels(&self) -> usize

Number of classes; labels are 0..num_labels.

pub fn insert(&mut self, data_point: DataPoint, feature_index: usize)

Appends an observation to column feature_index. The observation of instance 0 opens a new column.

pub fn set_num_label(&mut self, value: usize)

Sets the number of classes.

pub fn sort_features(&mut self)

Sorts every column by value.

pub fn compute_unique_feature_values(&mut self)

Assigns each observation the index of its value among the column’s distinct values. It may run before or after sort_features, and must run before any fit.

impl Dataset

pub fn from_rows( values: &[f64], labels: &[usize], n_features: usize, ) -> Result<Self, DatasetError>

Builds a dataset from a row-major value buffer and its labels.

This is the in-memory counterpart of DataReader::read_file, and takes numpy’s C-order (n_rows, n_features) layout directly.

The returned dataset is prepared: its columns are sorted and indexed.

Trait Implementations

impl Debug for Dataset

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for Dataset

fn default() -> Dataset

Returns the “default value” for a type.

impl Index<usize> for Dataset

type Output = Feature

The returned type after indexing.

fn index(&self, index: usize) -> &Self::Output

Performs the indexing (container[index]) operation.

impl IndexMut<usize> for Dataset

fn index_mut(&mut self, index: usize) -> &mut Feature

Performs the mutable indexing (container[index]) operation.

impl<'a> IntoIterator for &'a Dataset

type Item = &'a Feature

The type of the elements being iterated over.

type IntoIter = Iter<'a, Feature>

Which kind of iterator are we turning this into?

fn into_iter(self) -> Self::IntoIter

Creates an iterator from a value.