vaderSentimentScores — Score tokenized documents with VADER-style sentiment rules.

vaderSentimentScores(documents) returns a compound sentiment score for each English tokenizedDocument. [compoundScores,positiveScores,negativeScores,neutralScores] = vaderSentimentScores(___) also returns positive, negative, and neutral proportions.

Syntax

compoundScores = vaderSentimentScores(documents)
compoundScores = vaderSentimentScores(documents,Name,Value)
[compoundScores,positiveScores,negativeScores,neutralScores] = vaderSentimentScores(___)

Inputs

Returns

Returned values from vaderSentimentScores depend on how many outputs the caller requests.

Errors

How vaderSentimentScores works

• documents must be a RunMat tokenizedDocument compatibility object with English language metadata.
• Compound scores are normalized to the range [-1, 1] from token sentiment values after applying booster, dampener, negation, capitalization, and punctuation modifiers.
• SentimentLexicon accepts a RunMat table with lowercase string Token values and numeric SentimentScore values in the documented range [-4, 4].
• Boosters and Dampeners accept string, character, cell, or string-array inputs. Column vectors are treated as word lists; multi-column string arrays are treated as row-wise n-grams with empty padding ignored.
• Negations accepts string, character, cell, or string-array word lists.
• RunMat ships a compact built-in English compatibility lexicon and modifier lists for common scripts and examples. For domain-specific or full-coverage scoring, pass an explicit SentimentLexicon table.
• The default lexicon is not yet the complete MATLAB/VADER lexicon, and exact MATLAB VADER scoring parity remains tracked by the broader Text Analytics compatibility issue.

GPU memory and residency

vaderSentimentScores operates on host tokenizedDocument objects and has no provider-resident kernel.

Examples

Score Tokenized Documents

documents = tokenizedDocument(["The book was VERY good!!!!"; "The book was not very good."]);
compoundScores = vaderSentimentScores(documents)

Expected output:

`compoundScores` is a numeric column vector with one score per document.

Return Sentiment Proportions

[compoundScores,positiveScores,negativeScores,neutralScores] = vaderSentimentScores(documents)

Expected output:

Each output is a numeric column vector with one value per document.

Use a Custom Lexicon

tbl = table(["innovative"; "risk"], [4; -3], "VariableNames", ["Token", "SentimentScore"]);
compoundScores = vaderSentimentScores(documents, "SentimentLexicon", tbl)

Expected output:

The custom table replaces the built-in compact lexicon.

Using vaderSentimentScores with coding agents

Open a RunMat example with live inputs, then ask the agent to explain how vaderSentimentScores changes the result.

FAQ

Related Strings functions

Open-source implementation

Unlike proprietary runtimes, every RunMat function is open-source. Read exactly how vaderSentimentScores is executed, line by line, in Rust.

• View the source for vaderSentimentScores in Rust on GitHub
• Learn how the RunMat runtime works
• Found a bug? Open an issue with a minimal reproduction.

About RunMat

RunMat is an open-source runtime that executes MATLAB-syntax code blazing on any GPU. It is licensed under the Apache 2.0 license.

• RunMat automatically optimizes your math for GPU execution on Apple, Nvidia, and AMD hardware. No code changes needed. Simulations that took hours now take minutes.
• Start running code in seconds. RunMat runs in the browser, on the desktop, or from the CLI. No license server, no IT ticket.

Getting started · Benchmarks · Pricing