Clarity Platform

https://docs.fusion.metaspectral.com
Admin
Clarity is a cloud-based platform for processing, analyzing, and building machine learning models on...
Tokens:38,209
Snippets:420
Trust Score:8.8
Update:2 weeks ago
Show doc for...
Tokens:
Raw
### Install fusion_sdk wheel in a virtual environment

Source: https://docs.fusion.metaspectral.com/sdk

Install the fusion_sdk wheel in a Python 3.10+ virtual environment. On Windows, activate with `.venv\Scripts\activate` instead of `source .venv/bin/activate`.

```bash
python -m venv .venv  
source .venv/bin/activate  
pip install /path/to/fusion_sdk-<version>-py3-none-any.whl  
```

--------------------------------

### Get dataset version and download artifact

Source: https://docs.fusion.metaspectral.com/sdk-dataset-versions

Retrieves a dataset version by ID, prints its image splits and class index map, inspects the first label's class name and mask shape, then downloads the dataset artifact to a local folder. The download creates a subdirectory named after the version ID.

```python
dataset_version = sdk.dataset_version.get(dataset_version_id=2929)  
  
print("Train images:", dataset_version.train_image_ids)  
print("Validation images:", dataset_version.validation_image_ids)  
print("Test images:", dataset_version.test_image_ids)  
print("Class index map:", dataset_version.class_id_to_index_map)  
  
first_label = dataset_version.labels[0]  
print(first_label.class_name, first_label.mask.shape)  
  
dataset_version.download(folder_path="/local/downloads/datasets")  
```

--------------------------------

### SDK Folder: Get folder and download images

Source: https://docs.fusion.metaspectral.com/sdk-folders

Retrieves a folder by ID, iterates over its direct images, prints the count of all images in the tree, and downloads them recursively to a local path. Set recursive=False to download only images directly in the folder.

```python
folder = sdk.folder.get(folder_id=19)  
  
for image in folder.images:  
    print(image.id, image.name)  
  
print("Images in this tree:", len(folder.recursive_images))  
folder.download(folder_path="/local/downloads", recursive=True)  
```

--------------------------------

### Create and inspect a class collection with sdk.class_collection

Source: https://docs.fusion.metaspectral.com/sdk-class-collections

Creates a class collection named "Land cover", adds two classes, then retrieves the collection by ID and prints each class's ID and name. Requires an initialized SDK instance with the class_collection module.

```python
collection = sdk.class_collection.create_class_collection(name="Land cover")  
  
collection.create_class(name="Water")  
collection.create_class(name="Vegetation")  
  
saved = sdk.class_collection.get(id=collection.id)  
for cls in saved.get_classes():  
    print(cls.id, cls.name)  
```

--------------------------------

### Create a FusionSDKV2 client with service token

Source: https://docs.fusion.metaspectral.com/sdk

Creates a production client using the `CLARITY_SERVICE_TOKEN` environment variable. `FusionSDKV2` requires `service_token` and accepts optional `environment` and `api_url` arguments. Do not embed a service token in source control.

```python
import os  
  
from fusion_sdk.implementations_v2.common.enums import SDKEnvironmentEnum  
from fusion_sdk.implementations_v2.FusionSDKV2 import FusionSDKV2  
  
sdk = FusionSDKV2(  
    service_token=os.environ["CLARITY_SERVICE_TOKEN"],  
    environment=SDKEnvironmentEnum.PROD,  
)  
```

--------------------------------

### Upload, label, and download an EnMAP image with fusion_sdk

Source: https://docs.fusion.metaspectral.com/sdk-images

Upload an EnMAP HSI file bundle with explicit processing level and GeoTIFF source, then attach shapefile labels and download a processed layer. Pass every required file in file_paths; geotiff_source is optional for non-GeoTIFF imports. The processing_level labels values but does not convert or calibrate them, and a returned resource does not confirm background import processing has finished.

```python
from fusion_sdk.models.data_source_enum import DataSourceEnum  
from fusion_sdk.models.hsi_processing_level_enum import HsiProcessingLevelEnum  
  
image = sdk.image.upload(  
    file_paths=[  
        "/path/to/METADATA.XML",  
        "/path/to/SPECTRAL_IMAGE.TIF",  
    ],  
    name="imported_enmap",  
    folder_id=19,  
    processing_level=HsiProcessingLevelEnum.RADIANCE,  
    geotiff_source=DataSourceEnum.ENMAP,  
)  
  
# Continue after import processing has finished and the input layer is ready.  
image.upload_shapefiles(  
    workspace_id=8,  
    shapefile_paths=[  
        "/path/to/polygon.shp",  
        "/path/to/polygon.shx",  
        "/path/to/polygon.dbf",  
        "/path/to/polygon.prj",  
    ],  
)  
  
image.download(  
    folder_path="/local/downloads",  
    correction_level=HsiProcessingLevelEnum.RADIANCE,  
)  
```

--------------------------------

### Fetch, download, and query model versions with sdk.model_version

Source: https://docs.fusion.metaspectral.com/sdk-model-versions

Shows how to fetch a model version by ID, download its artifacts, read thresholded inference arrays with CLASSIFY mode, and query model versions by a timezone-aware date range.

```python
from datetime import datetime, timezone  
  
from fusion_sdk.models.thresholding_mode_enum import ThresholdingModeEnum  
  
# Fetch and download one model version.  
model = sdk.model_version.get(model_version_id=38)  
model.download(folder_path="/local/downloads/models")  
  
# Read thresholded inference arrays.  
for inference_result in model.inference_results:  
    classified = inference_result.get_data(  
        threshold_start=0.2,  
        threshold_end=0.8,  
        threshold_mode=ThresholdingModeEnum.CLASSIFY,  
    )  
    print(inference_result.image_id, classified.shape)  
  
# Use an explicit timezone for a date-range query.  
recent = sdk.model_version.get_by_date_range(  
    model_trained_after=datetime(2025, 1, 1, tzinfo=timezone.utc),  
    model_trained_before=datetime(2026, 1, 1, tzinfo=timezone.utc),  
    exclude_draft=True,  
)  

```

--------------------------------

### sdk.class_collection Module

Source: https://docs.fusion.metaspectral.com/sdk-class-collections

Documentation for the sdk.class_collection module methods and the ClassCollectionResource methods.

```APIDOC
## sdk.class_collection.get(id: int)

### Description
Returns the class collection with the specified ID. Raises `ValueError` when the ID is not found.

### Method
`get`

### Parameters
- **id** (int) - Required - The ID of the class collection to retrieve.

### Returns
- **ClassCollectionResource** - The class collection with the given ID, exposing `id` and `name` attributes.

### Errors
- `ValueError` - Raised when the ID is not found.

## sdk.class_collection.create_class_collection(name: str)

### Description
Creates and returns a new class collection with the specified name.

### Method
`create_class_collection`

### Parameters
- **name** (str) - Required - The name of the class collection to create.

### Returns
- **ClassCollectionResource** - The newly created class collection, exposing `id` and `name` attributes.

## ClassCollectionResource.create_class(name: str)

### Description
Creates and returns a class within the class collection.

### Method
`create_class`

### Parameters
- **name** (str) - Required - The name of the class to create.

### Returns
- **Class** - The newly created class.

## ClassCollectionResource.get_classes()

### Description
Returns the classes belonging to the class collection.

### Method
`get_classes`

### Parameters
None

### Returns
- **list[Class]** - The collection's classes.
```

--------------------------------

### InferenceResultResource.download

Source: https://docs.fusion.metaspectral.com/sdk-model-versions

Downloads and extracts the inference raster package to a local folder.

```APIDOC
## InferenceResultResource.download

### Description
Downloads and extracts the inference raster package to a specified local folder, with optional post-processing and view type.

### Method
GET

### Endpoint
InferenceResultResource.download

### Parameters
#### Path Parameters
- **folder_path** (string) - Required - The local directory where the inference raster package will be downloaded and extracted.

#### Query Parameters
- **post_processing_method** (enum) - Optional - The post-processing method to apply. Defaults to NONE.
- **view_type** (enum) - Optional - The view type for the raster. Defaults to BASE_MAP.

### Request Example
```python
inference_result.download(
    folder_path="/local/downloads",
    post_processing_method=NONE,
    view_type=BASE_MAP,
)
```

### Response
#### Success Response (200)
- **None** - The method downloads and extracts the inference raster package to the specified folder.

#### Response Example
```json
{}
```
```
Tokens:
Raw