| # Complete the Missing Response Capability |
|
|
| You are given a reduced source snapshot from a TERPSICHORE-style ideal MHD |
| stability workflow. |
|
|
| The current implementation already reproduces a simpler baseline regime. |
| However, the included public cases introduce richer response settings with |
| multiple coupled mode groups and radially localized transport behavior. In the |
| current source snapshot, the reproduced downstream behavior still looks too |
| close to the baseline regime across those settings. |
|
|
| Inspect the source and run: |
|
|
| ```bash |
| python reproduce.py |
| ``` |
|
|
| The reproduction writes the observed baseline and target quantities under |
| `outputs/`. The command returns nonzero if those observations are empty, |
| incomplete, or non-finite; it does not apply verifier-only scientific |
| acceptance windows. |
|
|
| Your task is to complete the implementation so that these richer settings are |
| handled as genuinely supported scientific regimes. |
|
|
| The intended capability chain is: |
|
|
| ```text |
| regime handling -> radial response profiles -> sideband/transport representation -> projection and pair support -> downstream blocks -> energy diagnostics |
| ``` |
|
|
| Do not hard-code: |
|
|
| - the included public case; |
| - a fixed list of expected numbers; |
| - a single mode table; |
| - one pair-support pattern; |
| - one coefficient tensor or one matrix block. |
|
|
| Do not bypass the scientific data path by writing directly into downstream |
| matrix or energy outputs. The implementation should support the general |
| workflow represented by the supplied task. |
|
|
| Internet access is not available during evaluation. |
|
|
