# Combine uncertainties by maximum

## Summary
Select the maximum absolute uncertainty among configured source keys.

## Metadata
- **Module ID:** CombineUncertaintiesMax
- **Module path:** /home/runner/work/MoDaCor/MoDaCor/src/modacor/modules/base_modules/combine_uncertainties_max.py
- **Module version:** 20260120.1
- **Keywords:** uncertainties, combine, maximum, propagation

## Required data keys
- signal

## Modifies
- **signal**: uncertainties

## Required arguments
- combinations

## Default configuration
```json
{
  "combinations": {},
  "drop_source_keys": false,
  "ignore_missing": false,
  "target_basedata_key": "signal"
}
```

## Argument specification
| Argument | Type | Required | Default | Description |
|---|---|---|---|---|
| `combinations` | dict | Yes | {} | Mapping of output uncertainty key to an iterable of source keys to combine. |
| `drop_source_keys` | bool | No | False | Remove source uncertainty keys after combination (default: False). |
| `ignore_missing` | bool | No | False | If True, missing source keys are ignored. If all listed keys are missing, the combination is skipped. |
| `target_basedata_key` | str | No | signal | Name of the BaseData entry within each DataBundle to modify (default: 'signal'). |

## References
DOI 10.1088/0953-8984/25/38/383201

## Notes
Useful when systematic contributions must be bounded by the most conservative estimate, mirroring needs in certain MOUSE workflows.