Options for Pattern Model loading and training.
More...
#include <patternmodel.h>
Options for Pattern Model loading and training.
This class defines all kinds of parameters that can be set for loading and training Pattern Models, it is passed to various constructors and methods.
| PatternModelOptions::PatternModelOptions |
( |
| ) |
|
|
inline |
Initialise with default values. All members are public and can be set explicitly..
| bool PatternModelOptions::DEBUG |
Output extra debug information.
| bool PatternModelOptions::DOPATTERNPERLINE |
Assume each line contains one integral pattern, rather than actively extracting all subpatterns on a line (default: false)
| bool PatternModelOptions::DOREMOVEFLEXGRAMS |
Remove flexgrams from the model upon loading it.
| bool PatternModelOptions::DOREMOVEINDEX |
Do not load index information (for indexed models), loads just the patterns without any counts.
| bool PatternModelOptions::DOREMOVENGRAMS |
Remove n-grams from the model upon loading it.
| bool PatternModelOptions::DOREMOVESKIPGRAMS |
Remove skip-grams from the model upon loading it.
| bool PatternModelOptions::DORESET |
sets all counts to zero upon loading, clears indices
| bool PatternModelOptions::DOREVERSEINDEX |
Obsolete now, only here for backward-compatibility with v1.
| bool PatternModelOptions::DOSKIPGRAMS |
Load/extract skipgrams? (default: false)
| bool PatternModelOptions::DOSKIPGRAMS_EXHAUSTIVE |
Load/extract skipgrams in an exhaustive fashion? More memory intensive, but the only options for unindexed models (default: false)
| int PatternModelOptions::MAXBACKOFFLENGTH |
Counting n-grams is done iteratively for each increasing n. (default: MAXLENGTH) For each n, presence of sub-ngrams in n-1 is checked. This values defines a maximum length for this back-off check. In combination with MINLENGTH, this allows earlier pruning and conserves memory.
| int PatternModelOptions::MAXLENGTH |
The maximum length of patterns to be loaded/extracted, inclusive (in words/tokens) (default: 100)
| int PatternModelOptions::MAXSKIPS |
Maximum skips per skipgram.
| int PatternModelOptions::MINLENGTH |
The minimum length of patterns to be loaded/extracted (in words/tokens) (default: 1)
| int PatternModelOptions::MINSKIPTYPES |
Minimum required amount of distinct patterns that can fit in a gap of a skipgram for the skipgram to be included (default: 2)
| int PatternModelOptions::MINTOKENS |
The occurrence threshold, minimum amount of occurrences for a pattern to be included in a model Defaults to 2 for building, to 1 for loading.
| int PatternModelOptions::MINTOKENS_SKIPGRAMS |
The occurrence threshold for skipgrams, minimum amount of occurrences for a pattern to be included in a model. Defaults to the same value as MINTOKENS. Only used if DOSKIPGRAMS or DO_SKIPGRAMS_EXHAUSTIVE is set to true
| int PatternModelOptions::MINTOKENS_UNIGRAMS |
The occurrence threshold for unigrams, unigrams must occur at least this many times for higher-order ngram/skipgram to be included in a model Defaults to the same value as MINTOKENS Only has an effect if MINTOKENS_UNIGRAMS > MINTOKENS.
| int PatternModelOptions::PRUNENONSUBSUMED |
| bool PatternModelOptions::QUIET |
The documentation for this class was generated from the following file:
1.8.10
