Evaluation

evaluation Module

Module for the evaluation of sequence alignment accuracy.

yalign.evaluation.F_score(xs, ys, beta=0.01)

Returns the F score described here: http://en.wikipedia.org/wiki/F1_score for list xs against to the list ys.

Make beta smaller to give more weight to the precision.

yalign.evaluation.alignment_percentage(document_a, document_b, model)

Returns the percentage of alignments of document_a and document_b using the model provided.

• document_a and document_b are two lists of Sentences to align.
• model can be a YalignModel or a path to a yalign model.

yalign.evaluation.classifier_precision(document_a, document_b, model)

Runs a ten-fold validation on the classifier and returns a value between 0 and 100. Higher is better.

yalign.evaluation.correlation(classifier, dataset=None)

Calculates the correlation of the attributes of a classifier.

For more information see:

• http://en.wikipedia.org/wiki/Correlation_and_dependence

yalign.evaluation.evaluate(parallel_corpus, model, N=100)

Returns statistics for N document alignment trials. The documents are generated from the parallel corpus.

• parallel_corpus: A file object
• model: A YalignModel
• N: Number of trials

yalign.evaluation.precision(xs, ys)

Precision of list xs to list ys.

yalign.evaluation.recall(xs, ys)

Recall of list xs to list ys.