SpatialLeaveAsideOut¶
-
class
museotoolbox.cross_validation.
SpatialLeaveAsideOut
(distance_matrix, valid_size=0.5, n_repeats=False, random_state=False, verbose=False)[source]¶ Generate a Cross-Validation using the farthest distance between the training and validation samples.
- Parameters
distance_matrix (numpy.ndarray, shape [n_samples, n_samples]) – Array got from function samplingMethods.getdistance_matrixForDistanceCV(inRaster,inVector)
valid_size (float, default 0.5.) – The percentage of validaton to keep : from 0 to 1.
n_repeats (int or bool, optional (default=False)) – If False, n_repeats is 1/valid_size (default : 1/0.5 = 2) If int : will iterate the number of times given in n_repeats.
random_state (integer or None, optional (default=None)) – If int, random_state is the seed used by the random number generator; If None, the random number generator is created with
time.time()
.verbose (integer or False, optional (default=False)) – Controls the verbosity: the higher the value is, the more the messages are detailed.
References
See “Combining ensemble modeling and remote sensing for mapping individual tree species at high spatial resolution” : https://doi.org/10.1016/j.foreco.2013.07.059.
Manage cross-validation methods to generate the duo valid/train samples.
Methods
__init__
(distance_matrix[, valid_size, …])Manage cross-validation methods to generate the duo valid/train samples.
get_n_splits
([X, y, groups])Returns the number of splitting iterations in the cross-validator.
save_to_vector
(vector, field[, group, …])Save to vector files each fold from the cross-validation.
split
(X, y[, groups])Split the vector/array according to y and groups.