# ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (3,) + inhomogeneous part.

- **ID:** `python/numpy-visible-deprecation-ragged`
- **Domain:** python
- **Category:** data_error
- **Verification:** ai_generated
- **Fix Rate:** 80%

## Root Cause

np.array() called on a list of lists with differing lengths. Since NumPy 1.24 this raises instead of creating an object array.

## Version Compatibility

| Version | Status | Introduced | Deprecated |
|---------|--------|------------|------------|
| 3.x | active | — | — |

## Workarounds

1. **** (95% success)
   ```
   np.array(data, dtype=object) to explicitly request an object array
   ```
2. **** (85% success)
   ```
   Pad sequences first: from itertools import zip_longest; np.array(list(zip_longest(*data, fillvalue=0))).T
   ```

## Dead Ends

- **** — Still cannot broadcast ragged data (90% fail)
- **** — np.stack requires identical shapes too (85% fail)