🪢 Open source agent evals & observability: Trace, evaluate, and improve LLM applications with one open platform.
TypeScript analyticsautogenevaluationlangchainlarge-language-modelsllama-indexllmllm-evaluationllm-observabilityllmopsmonitoringobservabilityopen-sourceopenaiplaygroundprompt-engineeringprompt-managementself-hostedycombinator
import { useMemo } from "react";
import { constructDatasetRunAggregateColumns } from "@/src/features/datasets/components/DatasetRunAggregateColumnHelpers";
import { api } from "@/src/utils/api";
import {
datasetRunItemsTableColsWithOptions,
type FilterState,
} from "@langfuse/shared";
import { scoreFilters } from "@/src/features/scores/lib/scoreColumns";
export function useDatasetRunAggregateColumns({
projectId,
runIds,
datasetId,
updateRunFilters,
getFiltersForRun,
}: {
projectId: string;
runIds: string[];
datasetId: string;
updateRunFilters: (runId: string, filters: FilterState) => void;
getFiltersForRun: (runId: string) => FilterState;
}) {
const datasetRunItemsFilterOptionsResponse =
api.datasets.runItemFilterOptions.useQuery({
projectId,
datasetId,
datasetRunIds: runIds,
});
const runsData = api.datasets.baseRunDataByDatasetId.useQuery(
{
projectId,
datasetId,
},
{
refetchOnWindowFocus: false,
refetchOnMount: false,
},
);
const scoreKeysAndProps = api.scores.getScoreColumns.useQuery(
{
projectId,
filter: scoreFilters.forDatasetRunItems({
datasetRunIds: runIds,
datasetId,
}),
},
{
enabled: runIds.length > 0,
},
);
const datasetRunItemsFilterOptions =
datasetRunItemsFilterOptionsResponse.data;
const datasetColumns = useMemo(() => {
return datasetRunItemsTableColsWithOptions(datasetRunItemsFilterOptions);
}, [datasetRunItemsFilterOptions]);
const runAggregateColumnProps = runIds.map((runId) => {
const runNameAndMetadata = runsData.data?.find((name) => name.id === runId);
return {
name: runNameAndMetadata?.name ?? `run-${runId}`,
id: runId,
description: runNameAndMetadata?.description ?? undefined,
createdAt: runNameAndMetadata?.createdAt,
};
});
const runAggregateColumns = useMemo(() => {
return constructDatasetRunAggregateColumns({
runAggregateColumnProps,
projectId,
datasetColumns,
serverScoreColumns: scoreKeysAndProps.data?.scoreColumns,
updateRunFilters,
getFiltersForRun,
});
}, [
runAggregateColumnProps,
projectId,
datasetColumns,
scoreKeysAndProps.data?.scoreColumns,
updateRunFilters,
getFiltersForRun,
]);
return {
runAggregateColumns,
isLoading: scoreKeysAndProps.isLoading,
};
}