vgherard / sbo

Utilities for training and evaluating text predictors based on Stupid Back-off N-gram models.

vgherard/sbo Issues

Pruning the dictionary
Closed 2 years ago4
Formula `dict` argument of `kgram_freqs()` works only with literals
Updated 4 years ago
Define S3 class for output of `eval_sbo_predictor()`
Updated 4 years ago
Implement PrefixCompletion C++ class through std::unordered_map
Updated 4 years ago
Implement next-word prediction from out-of-memory kgram frequency table
Updated 4 years ago
Vectorizing eval_sbo_preds()
Updated 4 years ago
twitter_test overlaps twitter_train...
Closed 4 years ago
PREPERROR: ‘unordered_map’ is not a member of ‘std’
Closed 4 years ago
NOTE: no visible global function definition for plot
Closed 4 years ago1
FAILURE (test-eval_sbo_predictor.R:19:9): output has the correct structure
Closed 4 years ago1
Eliminate global variables using the `.data$` pronoun instead
Closed 4 years ago
Properly document `plot()` method for `word_coverage` class
Closed 4 years ago
build_sbo_preds() directly from training corpus
Closed 4 years ago
Add build_dictionary() function
Closed 4 years ago
Utilities to compute coverage fraction of k-grams in a corpus
Closed 4 years ago
Document `filtered` argument of `build_sbo_preds()`
Closed 4 years ago
Huge memory allocations from `predict.sbo_preds`
Closed 4 years ago1
Changing data structure for `sbo_preds`
Closed 4 years ago
Using Run Length Encoding in `sbo_preds` objects.
Closed 4 years ago
Clean up Depends and Imports statements in DESCRIPTION
Closed 4 years ago
Add package documentation entry
Closed 4 years ago
Rethinking structure of `kgram_freqs` and `sbo_preds` object
Closed 4 years ago1
`summary()` and `str()` methods for `sbo_preds` and `kgram_freqs` S3 classes
Closed 4 years ago1
Tests for `get_kgram_freqs_fast()` failing on `rhub::platforms()[12,1]` (i.e. "macos-highsierra-release-cran")
Closed 4 years ago
Implement reading from file of training corpus in `get_kgram_freqs_fast()`
Updated 4 years ago