LeaveCloseCompositionsOut

class mastml.data_splitters.LeaveCloseCompositionsOut(composition_df, dist_threshold=0.1, nn_kwargs=None, **kwargs)

Bases: BaseSplitter

Leave-P-out where you exclude materials with compositions close to those the test set

Computes the distance between the element fraction vectors. For example, the \(L_2\) distance between Al and Cu is \(\sqrt{2}\) and the \(L_1\) distance between Al and Al0.9Cu0.1 is 0.2.

Consequently, this splitter requires a list of compositions as the input to split rather than the features.

Attributes:

parallel_run: an attribute definining wheteher to run splits with all available computer cores

Args:

composition_df (pd.DataFrame): dataframe containing the vector of material compositions to analyze

dist_threshold (float): Entries must be farther than this distance to be included in the training set

nn_kwargs (dict): Keyword arguments for the scikit-learn NearestNeighbor class used to find nearest points

Methods Summary

get_n_splits([X, y, groups])	Returns the number of splitting iterations in the cross-validator
split(X[, y, groups])	Generate indices to split data into training and test set.

Methods Documentation

get_n_splits(X=None, y=None, groups=None)

Returns the number of splitting iterations in the cross-validator

split(X, y=None, groups=None)

Generate indices to split data into training and test set.

Parameters

Xarray-like of shape (n_samples, n_features)

Training data, where n_samples is the number of samples and n_features is the number of features.

yarray-like of shape (n_samples,)

The target variable for supervised learning problems.

groupsarray-like of shape (n_samples,), default=None

Group labels for the samples used while splitting the dataset into train/test set.

Yields

trainndarray

The training set indices for that split.

testndarray

The testing set indices for that split.