qubitclient-nnscope

Files (1)

• SKILL.md

SKILL.md (first lines)

## Overview

NNScope is a neural network-based quantum spectrum analysis module that provides advanced curve segmentation, peak detection, and parameter extraction tasks. Unlike traditional scope fitting tasks, NNScope uses deep learning models for more robust analysis on complex spectral data.

## API Reference

### Client Initialization

```python
from qubitclient import QubitNNScopeClient, NNTaskName
from qubitclient.nnscope.nnscope_api.curve.curve_type import CurveType

client = QubitNNScopeClient()
```

### Task Names (NNTaskName)

| Task | Description |
|------|-------------|
| `S21PEAK` | S21 peak detection with confidence scoring |
| `S21PEAKMULTI` | Multi-peak detection across full frequency range |
| `SPECTRUM` | Spectrum peak region analysis (start, end, width, confidence) |
| `SPECTRUM2D` | 2D spectrum curve segmentation (supports COSINE, POLY curve types) |
| `S21VSFLUX` | S21 vs Flux parameter curve segmentation |
| `POWERSHIFT` | Power shift curve segmentation |

### Data Input Formats

The input data should be a dictionary or list of dictionaries with the following structure:

```python
dict_list = [{
    "image": {
        "Q0": [x_data, y_data],  # or [x_data, y_data, z_data] for 2D
        "Q1": [x_data, y_data],
    }
}, ...]  # One or more data items
```

**Note:** Multiple data items are supported - pass a list of dictionaries. The output `results` list has one entry per input item, in the same order.

---

#### S21PEAK

**Input:**
```python
{
    "image": {
        "Q0": [x_array, amp_array, phi_array],  # tuple, length=3
        "Q1": [x_array, amp_array, phi_array],
    }
}
```
- `x_array`: np.ndarray, shape=(N,), dtype=float64 - frequency data
- `amp_array`: np.ndarray, shape=(N,), dtype=float64 - amplitude data
- `phi_array`: np.ndarray, shape=(N,), dtype=float32 - phase data

**Output:**
```json
{
  "type": "s21peak",
  "results": [{
    "peaks": [[int, ...], [int, ...]],              // Peak indices per qubit
    "confs": [[float, ...], [float, ...]],          // Confidence scores per peak
    "freqs_list": [[float, ...], [float, ...]],     // Peak frequencies per qubit
    "status": "success" | "failed"
  }]
}
```

---

#### S21PEAKMULTI

Similar to S21PEAK but uses combined qubit keys (e.g., "Q0101") and does not include an `id` field.

**Input:**
```python
{
    "image": {
        "Q0101": [x_array, amp_array, phi_array],  # tuple, length=3
        "Q0202": [x_array, amp_array, phi_array],
    }
}
```
- `x_array`: np.ndarray, shape=(N,), dtype=float64 - frequency data
- `amp_array`: np.ndarray, shape=(N,), dtype=float32 - amplitude data
- `phi_array`: np.ndarray, shape=(N,), dtype=float32 - phase data

**Output:**
```json
{
  "type": "s21peakmulti",
  "results": [{
    "peaks": [[int, ...], [int, ...]],              // Peak indices per qubit
    "confs": [[float, ...], [float, ...]],          // Confidence scores per peak
    "freqs_list": [[float, ...], [float, ...]],     // Peak frequencies per qubit
    "status": "success" | "failed"
  }]
}
```

---

#### SPECTRUM

**Input:**
```python
{
    "image": {
        "Q0": [x_array, y_array],  // tuple, length=2
        "Q1": [x_array, y_array],
    }
}
```
- `x_array`: np.ndarray, shape=(N,), dtype=float64 - frequency data
- `y_array`: np.ndarray, shape=(N,), dtype=float32 - spectral amplitude

From yaqiangsun/QubitClient, MIT.

More from yaqiangsun/QubitClient