eval

package

v0.5.0 Latest Latest Go to latest Published: Sep 25, 2026 License: MIT Imports: 16 Imported by: 0

Details

Valid go.mod file
Redistributable license
Tagged version
Stable version
Learn more about best practices

Repository

github.com/diffpal/lintpal

Links

Open Source Insights

Documentation ¶

Overview ¶

Package eval evaluates frozen examples with typed, local answers. It does not call a provider and its counts do not measure live model accuracy.

Index ¶

Constants
Variables
type Case
type Corpus
- func Load(reader io.Reader) (Corpus, error)
type Outcome
type Summary
- func RunOffline(ctx context.Context, corpus Corpus) (Summary, []report.Report, error)

Constants ¶

View Source

const MaxCorpusBytes = 256 << 10

View Source

const SchemaVersion = "lintpal.eval.corpus.v1"

Variables ¶

View Source

var ErrInvalidCorpus = errors.New("invalid evaluation corpus")

Functions ¶

This section is empty.

Types ¶

type Case ¶

type Case struct {
	ID          string  `json:"id"`
	Path        string  `json:"path"`
	Line        int     `json:"line"`
	Source      string  `json:"source"`
	RuleID      string  `json:"rule_id"`
	Truth       string  `json:"truth"`
	Probability float64 `json:"probability"`
}

type Corpus ¶

type Corpus struct {
	SchemaVersion string `json:"schema_version"`
	Cases         []Case `json:"cases"`
}

func Load ¶

func Load(reader io.Reader) (Corpus, error)

type Outcome ¶

type Outcome struct {
	ID          string `json:"id"`
	RuleID      string `json:"rule_id"`
	Truth       string `json:"truth"`
	Predicted   bool   `json:"predicted"`
	Diagnostics int    `json:"diagnostics"`
}

type Summary ¶

type Summary struct {
	SchemaVersion string    `json:"schema_version"`
	CorpusVersion string    `json:"corpus_version"`
	Cases         []Outcome `json:"cases"`
	TruePositive  int       `json:"true_positive"`
	TrueNegative  int       `json:"true_negative"`
	FalsePositive int       `json:"false_positive"`
	FalseNegative int       `json:"false_negative"`
}

func RunOffline ¶

func RunOffline(ctx context.Context, corpus Corpus) (Summary, []report.Report, error)

RunOffline executes the real selection, decision, and report path against frozen examples. Probabilities are fixture inputs, not model predictions.

Source Files ¶

View all Source files

offline.go

?	: This menu
/	: Search site
f or F	: Jump to
y or Y	: Canonical URL