# Concatenate DataBundles

## Summary
Concatenate matching 1D BaseData entries, optionally sorting every entry together.

## Metadata
- **Import path:** `modacor.modules.base_modules.concatenate_databundles.ConcatenateDatabundles`
- **Source:** [`src/modacor/modules/base_modules/concatenate_databundles.py`](https://github.com/BAMresearch/MoDaCor/blob/main/src/modacor/modules/base_modules/concatenate_databundles.py)
- **Module ID:** ConcatenateDatabundles
- **Module version:** 20261002.1
- **Keywords:** concatenate, pool, curves, sort

## Required data keys
- _None_

## Modifies
- **configured data keys**: signal, uncertainties, weights, units, axes
- **source_index**: signal

## Required arguments
- with_processing_keys
- data_keys
- output_processing_key

## Default configuration
```json
{
  "alignment_key": null,
  "data_keys": [
    "signal",
    "Q"
  ],
  "descending": false,
  "output_processing_key": "concatenated",
  "sort_by": null,
  "source_index_key": "source_index",
  "source_position_key": null,
  "uncertainty_key_policy": "require_matching",
  "with_processing_keys": null
}
```

## Argument specification
| Argument | Type | Required | Default | Dependency role | Description |
|---|---|---|---|---|---|
| `alignment_key` | str or NoneType | No | - | - | Optional configured data key whose values must match pointwise across all inputs after unit conversion. Useful when source_position_key will group aligned observations. |
| `data_keys` | list or str | Yes | ["signal", "Q"] | - | Matching one-dimensional BaseData entries to concatenate. |
| `descending` | bool | No | False | - | Sort descending instead of ascending when sort_by is set. |
| `output_processing_key` | str | Yes | concatenated | - | ProcessingData key receiving the pooled DataBundle. |
| `sort_by` | str or NoneType | No | - | - | Optional concatenated data key used for coordinated stable sorting. |
| `source_index_key` | str or NoneType | No | source_index | - | Optional output key recording each point's zero-based input-bundle index. |
| `source_position_key` | str or NoneType | No | - | - | Optional output key recording each point's zero-based position within its input bundle. For aligned inputs, this can be used directly as an IndexedAverager index map. |
| `uncertainty_key_policy` | str | No | require_matching | - | How differently named uncertainty components are handled: 'require_matching' rejects them; 'fill_zero' takes their union and fills absent components with zero. Use fill_zero only when an absent component means no contribution from that source. |
| `with_processing_keys` | list | Yes | - | - | Input DataBundle keys, in concatenation order. |

## Notes
Input order is preserved when sort_by is None. Units are converted to those of the first input. Uncertainty component names must match across inputs unless uncertainty_key_policy='fill_zero'. Under fill_zero, absence means zero uncertainty from that named source, not unknown uncertainty. Sorting is stable and is not required by IndexByCoordinate. A source-position index groups points by array position only; set alignment_key when the step should verify that a configured coordinate matches pointwise across inputs.