drizzle-team/drizzle-orm · error · Error

The weights for the Weighted Random feature must add up to…

Error message

The weights for the Weighted Random feature must add up to exactly 1. Please review your weights to ensure they total 1 before proceeding

What it means

The Weighted Random feature (getWeightedIndices and the WeightedRandomGenerator) requires the supplied weights to sum to exactly 1.0 so they represent a valid probability distribution. The library uses integer-scaled arithmetic (1e10 scale) to avoid floating-point drift, so the sum must be exactly 1, not approximately.

Solutions

Adjust the weights so they sum to exactly 1 (e.g. [0.3, 0.3, 0.4]).
Compute the last weight as 1 minus the sum of the others to guarantee the total: lastWeight = 1 - others.reduce((a,b)=>a+b,0).
Double-check for floating-point issues by rounding to a fixed number of decimals.

Example fix

// before
weights: [0.2, 0.3, 0.3]  // sums to 0.8
// after
weights: [0.2, 0.3, 0.5]  // sums to 1.0

Defensive patterns

Strategy: validation

Validate before calling

function validateWeights(weights: number[]): string | null {
  const scale = 1e10;
  const sum = weights.reduce((acc, w) => acc + Math.round(w * scale), 0) / scale;
  return sum === 1 ? null : `Weights sum to ${sum}, expected exactly 1.`;
}
// Auto-fix last weight:
function normalizeWeights(weights: number[]): number[] {
  const fixed = weights.slice(0, -1);
  const last = 1 - fixed.reduce((a, b) => a + b, 0);
  return [...fixed, Math.round(last * 1e10) / 1e10];
}

Prevention

Always compute the last weight as 1 minus the sum of the others.
Use a schema-validation step or unit test to assert weights sum to 1.
Avoid hand-typing many small decimal weights; prefer fractions that sum cleanly.

When it happens

Trigger: Passing weights like [0.3, 0.3, 0.3] (sums to 0.9) or [0.5, 0.6] (sums to 1.1) to a weighted generator or to the `.with` weighted-count option.

Common situations: Rounding errors when splitting weights by hand; adding a new weighted option without re-normalizing the rest; copy-paste that duplicates a weight.

Related errors

AI-assisted analysis of drizzle-team/drizzle-orm@b7862528fd (2026-08-03). Data as JSON: /api/errors/184ff9f8370dfd73. Report an issue: GitHub.

Appendix: source

Thrown at drizzle-seed/src/services/utils.ts:30

	return resultList;
};

const sumArray = (weights: number[]) => {
	const scale = 1e10;
	const scaledSum = weights.reduce((acc, currVal) => acc + Math.round(currVal * scale), 0);
	return scaledSum / scale;
};

/**
 * @param weights positive number in range [0, 1], that represents probabilities to choose index of array. Example: weights = [0.2, 0.8]
 * @param [accuracy=100] approximate number of elements in returning array
 * @returns Example: with weights = [0.2, 0.8] and accuracy = 10 returning array of indices gonna equal this: [0, 0, 1, 1, 1, 1, 1, 1, 1, 1]
 */
export const getWeightedIndices = (weights: number[], accuracy = 100) => {
	const weightsSum = sumArray(weights);
	if (weightsSum !== 1) {
		throw new Error(
			`The weights for the Weighted Random feature must add up to exactly 1. Please review your weights to ensure they total 1 before proceeding`,
		);
	}

	// const accuracy = 100;
	const weightedIndices: number[] = [];
	for (const [index, weight] of weights.entries()) {
		const ticketsNumb = Math.floor(weight * accuracy);
		weightedIndices.push(...Array.from<number>({ length: ticketsNumb }).fill(index));
	}

	return weightedIndices;
};

export const generateHashFromString = (s: string) => {
	let hash = 0;
	// p and m are prime numbers
	const p = 53;

View on GitHub (pinned to b7862528fd)