Implement covariance support in uncertainty calculations (GUM)

## Enhancement: Covariance in Uncertainty Calculations (GUM)

**Summary:**
Implement support for covariance in uncertainty calculations as described in the Guide to the Expression of Uncertainty in Measurement (GUM). The function should accept a standardized input format for the covariance matrix.

**Implementation plan:**
- **Step 1:** Develop covariance handling for standard uncertainty calculations.
- Add input validation and parsing for standardized covariance matrix format.
- Integrate into existing uncertainty propagation functions.
- Ensure compatibility with other relevant modules/classes.
- Provide usage/examples and tests for matrix input.
- **Step 2 (Later):** Extend covariance handling for Monte Carlo uncertainty calculations.
- Adapt input parsing for MC simulation routines.
- Ensure proper random sampling respecting provided covariance structure.
- Include relevant documentation and unit tests.

**Requirements:**
- The covariance matrix input should be standardized (e.g., numpy array, dictionary with structure, or other clear format—please decide and document the chosen standard).
- Follow GUM guidelines for propagation with correlated inputs.
- Provide in-code examples and references to GUM.

**Additional Notes:**
- First implementation should be for standard uncertainty calculations.
- Monte Carlo support is required later, and this issue can serve as the parent for a follow-up sub-issue.

The issue does not name files, tests, or entry points. Start by locating the existing standard uncertainty propagation functions and their input conventions. Done means choosing and documenting a covariance format, validating and parsing it, integrating GUM covariance propagation, and adding examples and unit tests; Monte Carlo support is explicitly deferred.

Written by the indexing model from the issue text.