Skip to content
Browse documentation
Intelligence / ATH documentation

ath grade

Grade an existing prediction against an existing result.

Usage

$ath grade <prediction-id> --result <result-id> --json

Arguments

<prediction-id>
Required stable prediction record ID.

What it does

Grades two records that already exist. It creates a stable prediction-result link when needed and saves deterministic arithmetic against exact revisions. It never logs a benchmark result.

Examples

Use ath records list to find prediction and benchmark-result IDs. Preview the exact link and grade, then repeat with --yes. A prediction error analysis is a separate authored record written through structured ath log.

$ath records list --type prediction
$ath records list --type benchmark_result
$ath grade <prediction-id> --result <result-id> --yes

Options

--result <record-id>
stable benchmark-result record ID
--calculation-version <version>
grade calculation version; defaults to prediction-grade-v1
--file <path>
change a file other than the one in this folder
-y, --yes
write it without asking — for scripts, and when you are sure
--dry-run
show what would be written, and write nothing
--json
structured preview; never confirms the write

Related commands

Up nextath backtest →