chroma-core/chroma · error · ValueError

SparseVector values must be numbers, got {type(val).__name__

Error message

SparseVector values must be numbers, got {type(val).__name__} at position {i}

What it means

SparseVector.__post_init__ requires every element of `values` to be an instance of Python `int` or `float`. Crucially, NumPy float scalars (np.float32, np.float64) are NOT subclasses of Python float, so numerically valid values pulled element-wise from a NumPy array still fail. None and strings fail as well. The message reports the offending type and position.

Source

Thrown at chromadb/base_types.py:73

            if len(self.labels) != len(self.indices):
                raise ValueError(
                    f"SparseVector labels must have the same length as indices and values, "
                    f"got {len(self.labels)} labels, {len(self.indices)} indices"
                )

        for i, idx in enumerate(self.indices):
            if not isinstance(idx, int):
                raise ValueError(
                    f"SparseVector indices must be integers, got {type(idx).__name__} at position {i}"
                )
            if idx < 0:
                raise ValueError(
                    f"SparseVector indices must be non-negative, got {idx} at position {i}"
                )

        for i, val in enumerate(self.values):
            if not isinstance(val, (int, float)):
                raise ValueError(
                    f"SparseVector values must be numbers, got {type(val).__name__} at position {i}"
                )

        # Validate indices are sorted in strictly ascending order
        if len(self.indices) > 1:
            for i in range(1, len(self.indices)):
                if self.indices[i] <= self.indices[i - 1]:
                    raise ValueError(
                        f"SparseVector indices must be sorted in strictly ascending order, "
                        f"found indices[{i}]={self.indices[i]} <= indices[{i-1}]={self.indices[i-1]}"
                    )

    def to_dict(self) -> Dict[str, Any]:
        """Serialize to transport format with type tag.

        Note: Uses 'tokens' as the wire format key name for compatibility
        with the protobuf schema, even though the Python attribute is 'labels'.
        """

View on GitHub (pinned to aecdd12c8a)

Solutions

1. Convert before constructing: values = [float(v) for v in values] (float() accepts int, np.float32/64).
2. Call .tolist() on the NumPy array holding the values to get native floats.
3. Sanitize upstream: replace None/'' with 0.0 or drop the (index, value) pair before building the vector.

Example fix

// before
vals = [tfidf_matrix[0, j] for j in cols]  # np.float32 elements
sv = SparseVector(indices=cols, values=vals)  # ValueError: got float32

// after
vals = [float(tfidf_matrix[0, j]) for j in cols]
sv = SparseVector(indices=cols, values=vals)

# Normalize values before constructing SparseVector
values = [float(v) for v in values]  # np.float32/np.float64/int -> float

assert all(isinstance(v, (int, float)) for v in values)
sv = SparseVector(indices=indices, values=values)

def is_numeric_value_list(xs: object) -> bool:
    """Matches Chroma's check: isinstance(v, (int, float)) for every element."""
    return isinstance(xs, list) and all(isinstance(v, (int, float)) for v in xs)

try:
    sv = SparseVector(indices=indices, values=values)
except ValueError as e:
    raise ValueError(f'invalid sparse values for doc {doc_id}: {e}') from e

When it happens

Trigger: values taken element-wise from a NumPy array (arr[i] gives np.float32/np.float64); TfidfTransformer/sklearn outputs left as NumPy scalars; dicts parsed from JSON containing null or string numbers ('0.5') in the values slot.

Common situations: Sparse values from TF-IDF or BM25 pipelines that return NumPy matrices; values round-tripped through pandas; optional weights serialized as null/empty-string and passed through unmodified.

Related errors

• SparseVector indices must be integers, got {type(idx).__name
• Expected embeddings to be a list of floats or ints, a list o
• Sparse vector index must be created on a specific key. Pleas
• Deleting sparse vector index is not currently supported.
• limit must be a non-negative integer

AI-assisted analysis of chroma-core/chroma@aecdd12c8a (2026-08-16). Data as JSON: /api/errors/c7afe769c0ead9ca. Report an issue: GitHub.