Optimize lib for large datasets

**Problem**
When working with config, the error messages could be more descriptive.

**Suggestion**
Improve error handling and add detailed error messages.

**Context**
This would help with debugging and maintenance.

The issue mentions optimizing the library for large datasets and making config errors more descriptive, but names no files or tests. Start by locating the config handling and library entry points, then clarify which large-dataset behavior and error cases are in scope. Done should mean an agreed scope, improved diagnostic messages, and evidence that the relevant dataset path is optimized.

Written by the indexing model from the issue text.