lingdist - Fast Linguistic Distance and Alignment Computation
A fast linguistic distance and alignment computation
package. It implements generalized edit distance, Pointwise
Mutual Information (PMI) distance, and Weighted Jaccard
Distance (WJD). For generalized edit distance, the package
allows users to define custom cost for every symbol's
insertion, deletion, and substitution, and supports treating
character combinations as single symbols (e.g., IPA with
diacritics). It also provides detailed alignment information.
PMI distance automatically learns costs from data. WJD is
suitable for hierarchical categorical data with multiple forms.
All functions are implemented in 'C++' and distance matrix
computation is parallelized leveraging the 'RcppThread'
package.
