NEWS.md

In vspinu/mlvocab: Vocabulary and Corpus Preprocessing for Natural Language Pipelines

Version 0.1.2

• Number of threads can now be controlled with options("mlvocab.nthreads") or environment variable MLVOCAB_NTHREADS
• New argument keep_terms to prune_vocab

Version 0.1

• Automatically sort vocabs in decreasing term frequency order
• #3 Add support for character vector corpus.
• Add support for data.frame corpus
• New index generator function tix_df to produce flat indices as data.frames with provision for multi-keys for documents.
• Rename tixyz function into tix_xyz.
• Rename vocab_embed -> prune_embeddings, vocab_prune -> prune_vocab, vocab_update -> update_vocab.

Version 0.0.1

Initial Release

vspinu/mlvocab documentation built on June 11, 2021, 7:37 a.m.