Class: NILOD

NILOD(dbnullable)

new NILOD(dbnullable)

The NILOD model constructor.
Parameters:
Name Type Attributes Default Description
db string | sqlite.Database <nullable>
 null A filepath whereto load and save the NILOD model in the disk, or a SQLite database instance to use as the NILOD model's internal database.
Source:

Methods

(static) loadSave(save, filepathnon-null, encodingopt) → (nullable) {NILOD}

Loads a save retrieved by NILOD#getSave. Unfortunately this cannot be done without a file to store the save to (before opening it with the database object). Returns the loaded database. If there is already a file or directory at NILOD#loadSave.filepath, nothing will be done, and _null_ will be returned instead, to avoid damaging existing files.
Parameters:
Name Type Attributes Default Description
save string | Buffer The save, usually a base64-encoded string (as NILOD#getSave usually returns saves encoded as base64). To use another encoding, or buffers, please see the [encoding] section.
filepath string The path to a file that can be created, to be used as the database file. Unfortunately, SQLite databases may not be loaded in any other way.
encoding string <optional>
 base64 The encoding of the save. Defaults to _'base64'_.
Source:
Returns:
Type
NILOD

allEdgesBackward(nodeId) → {Promise.<Array.<Edge>>}

All edges that link _to_ the given node, from some other node in the model's graph.
Parameters:
Name Type Description
nodeId string The ID of the reference node in question.
Source:
Returns:
Type
Promise.<Array.<Edge>>

allEdgesForward(nodeId) → {Promise.<Array.<Edge>>}

All edges that link _from_ the given node to some other node in the model's graph.
Parameters:
Name Type Description
nodeId string The ID of the reference node in question.
Source:
Returns:
Type
Promise.<Array.<Edge>>

fetchEdges() → {Promise.<Array.<NILOD#Edge>>}

Fetchs a list of all the edges in the model's internal SQLite graph.
Source:
Returns:
Type
Promise.<Array.<NILOD#Edge>>

fetchHistory() → {Promise.<Array.<NILOD#OutputHistory>>}

Fetches a list of output history from the internal SQLite database.
Source:
Returns:
Type
Promise.<Array.<NILOD#OutputHistory>>

fetchNodes() → {Promise.<Array.<NILOD#Node>>}

Fetchs a list of all nodes from the internal SQL graph.
Source:
Returns:
Type
Promise.<Array.<NILOD#Node>>

findEdge(from, to) → {Promise.<?NILOD#Edge>}

Finds the particular edge between two nodes. Returns null if no edge connects both nodes.
Parameters:
Name Type Description
from string The ID of the node the edge stems *from*.
to string The ID of the node the edge ends in*to*.
Source:
Returns:
Type
Promise.<?NILOD#Edge>

findEdgesBetween(word1, word2) → {Promise.<Array.<NILOD#Edge>>}

Finds all edges that link a word to another, regardless of whichever nodes such words come from.
Parameters:
Name Type Description
word1 string The 'from' word.
word2 string The 'to' word.
Source:
Returns:
Type
Promise.<Array.<NILOD#Edge>>

generate(samples, config) → {NILOD#GenerationStatus}

Attempts to generate a good-looking sentence, by manufacturing N random sentences, and giving each one a rating, which in turn is predicted from previously generated sentences (output history).
Parameters:
Name Type Default Description
samples integer 30 The number of samples (random sentences) to generate and rate.
config object A configuration object that can hold properties for the config arguments of both NILOD#tryOnce (for per-sample generation) and NILOD#rateText (for sample rating). Please see the respective methods' _config_ arguments for more details on such properties.
Source:
Returns:
Type
NILOD#GenerationStatus

getNode(id) → {Promise.<?NILOD#Node>}

Retrieves a node by ID. Returns null if no node is found with the given ID.
Parameters:
Name Type Description
id string The ID of this node in particular.
Source:
Returns:
Type
Promise.<?NILOD#Node>

getSave()

Retrieves a base64 save of the SQLite database. Can be loaded using the NILOD#loadSave 'static' member.
Source:

rateText(text, config) → {Promise.<number>}

Predicts a rating to a given sentence or text, based on output history (and order in the tokenized text). This rating is the resolve value of the returned Promise.
Parameters:
Name Type Description
text string The text to rate.
config object The rating configuration
Properties:
Name Type Description
wordSimilarity function A word similarity function; a function that returns 1 for equal strings, and gradually lower numbers for gradually more different strings. Defaults to Levenshtein-based similarity.
orderDistanceGradient function A gradient function; computes the falloff of weight of a word, when compared with a word N tokens apart in the comparison sentence. Should be 1 for distance 0, and lower numbers (usually tending toward 0) for higher distances.
Source:
Returns:
Type
Promise.<number>

train(sentences) → {Promise}

Trains this model's Markov-like generator graph. If you specify a filename to its database, you may use an external sqlite database viewer, like the sqlite command-line utility, to view (and even edit) the graph's data (nodes and edges). Returns a Promise that is resolved when training is finished.
Parameters:
Name Type Description
sentences Array.<string> A list of sentences with the which to train the model's generator.
Source:
Returns:
Type
Promise

tryOnce(configopt) → {Promise.<NILOD#SentenceAttempt>}

Generates a random sentence based on the given Markov data. Returns a Promise that is resolved once the sentence is generated, returning the Attempt.
Parameters:
Name Type Attributes Description
config Object <optional>
 The settings for this attempt.
Properties
Name Type Attributes Default Description
query string <optional>
 The seed query for this sentence generation attempt. Random by default.
maxWords integer <optional>
 64 The maximum amount of words to be composed in the output.
maxChars integer <optional>
 512 The maximum amount of characters to form the output.
Source:
Returns:
Type
Promise.<NILOD#SentenceAttempt>

Type Definitions

Edge

A graph edge, arguably the most vital part of the internal NILOD graph.
Type:
  • Object
Properties:
Name Type Description
from string The graph node the edge stems *from*.
to string The graph node the edge ends in*to*.
fromWord string The particular word of the graph node the edge stems from.
toWord string The particular word of the graph node the edge ends onto.
separator string The string that delimits the fromWord and toWord, i.e. the separator between both nodes as per this edge.
count integer The number of occurences of this particular edge that were found during training. Used as a probabilistic measure during sentence generation.
Source:

GenerationStatus

Properties:
Name Type Description
best NILOD#OutputSample The predicted best sample, out of the samples generated.
worst NILOD#OutputSample The predicted worst sample, out of the samples generated. _Laughable._
bestResult string The predicted best sample's result sentence as a string.
scoreboard Array.<NILOD#OutputSample> A sorted list of samples, in descending order of rating.
Source:

Node

A node from a NILOD model graph. Nodes are simple structures that contain the bulk of word data. Those words are not compressed by a dictionary, but further redundancy is avoided via the usage of edges. Each item is a word. In NILOD, the array is always only 2 words long.
Type:
  • Array.<string>
Properties:
Name Type Description
id string The ID of this particular node.
Source:

OutputHistory

A single sentence generated by the NILOD model at any point in time.
Type:
  • Object
Properties:
Name Type Description
id string The ID of this sentence in particular. Usually a mnemonic formed by common words joined by hyphens.
data string The text that forms this sentence.
rating number The average rating of this sentence, usually ranging from -1 to 1.
rateCount number The number of ratings. Used to average them into the NILOD#OutputHistory.rating property.
seedID string The ID of the node that seeded this particular sentence.
Source:

OutputSample

Properties:
Name Type Description
rating number The rating of this sample.
attempt NILOD#SentenceAttempt The attempt wrapped by this sample.
Source:

SentenceAttempt

A sentence generated by the NILOD model.
Type:
  • Object
Properties:
Name Type Attributes Description
success boolean Whether this attempt was successful (i.e. no errors).
seedNode string <nullable>
 The ID of the node that was used as the seed (i.e. the starting point for this sentence).
seedText string <nullable>
 The string representation of the seed node (see above).
seedQuery string <nullable>
 The original query from the which the node was deduced (or randomly picked).
output string <nullable>
 The output of this query, or null if unsuccessful.
outputWords Array.<string> The words that were joined to compose the output property.
outputID string <nullable>
 The ID of this sentence generation, used when rating it.
reasonCode string <nullable>
 An upper-case, shorthand string that allows for quick mnemonic identification and comparison of error codes.
reasonMessage string <nullable>
 A human-readable reason describing the error.
reasonNumber Number An integer used to identify this particular reason.
applyRating function A method that allows rating this particular generated sentence. Takes a single argument (a rating, usually between -1 and 1), and returns a Promise that is resolved when the underlying SQL database is updated accordingly.
Source: