<div class="section-label"><span class="num">05</span> predict_psr</div>
<div class="section-title">Partial-sample regression through the <em>API</em>.</div>
`predict_psr` posts a [[API/Prediction Engine/PSR|PSR]] job and returns the prediction plus details.
Supports single-task, multi-y, and multi-theta shapes. Multi-y and multi-theta cannot be combined in one call.
## Signature
```text
predict_psr(
y, # ndarray, N × 1 or N × Q
X, # ndarray, N × K
theta, # ndarray, 1 × K or Q × K
options # PredictionOptions
) → yhat, yhat_details
```
**Call:** `predict_psr`
**Options:** [[Prediction Options#PredictionOptions|PredictionOptions]]
**REST:** `POST /psr`
## Arguments
| Name | Shape | Description |
| --- | --- | --- |
| <span class="mono">y</span> | N × 1, or N × Q for multi-y | Dependent variable(s) |
| <span class="mono">X</span> | N × K | Independent variables |
| <span class="mono">theta</span> | 1 × K, or Q × K for multi-theta | Circumstances |
| <span class="mono">options</span> | <span class="mono">PredictionOptions</span> | PSR configuration |
## Returns
| Name | Type | Description |
| --- | --- | --- |
| <span class="mono">yhat</span> | ndarray | Predicted outcome(s) |
| <span class="mono">yhat_details</span> | dict | Additional model and result details |
> [!warning]
> Raises <span class="mono">ValueError</span> if multi-y and multi-theta are specified together, or if <span class="mono">y</span>, <span class="mono">X</span>, and <span class="mono">theta</span> dimensions do not align.
## Related
- [[API/Python API Client/API Client/Overview|API Client]]
- [[predict_grid|predict_grid]] · [[predict_maxfit|predict_maxfit]]
- [[API/Prediction Engine/PSR|POST /psr]]