Common SDK methods
Cheatsheet for most common SDK methods for ontology management
Create new features and ontologies
When you are creating or importing features or ontologies for the first time, use the methods below. Visit our Python SDK guides for a complete list of supported feature types, or simply create a feature in the Schema tab and view its JSON.
from labelbox import OntologyBuilder, Tool, Classification, Option, Client
client = Client()
object_features = [
Tool(
tool=Tool.Type.BBOX,
name="regulatory-sign",
color="#ff0000",
classifications=[
Classification(
class_type=Classification.Type.RADIO,
name="Sign type",
options=[
Option(value="pedestrian", label="Pedestrian"),
Option(value="one_way", label="One way")
]
)
]
)
]
classification_features = [
Classification(
class_type=Classification.Type.CHECKLIST,
name="Quality Issues",
options=[
Option(value="blurry", label="Blurry"),
Option(value="distorted", label="Distorted")
]
)
]
ontology_builder = OntologyBuilder(
tools=object_features,
classifications=classification_features
)
ontology = client.create_ontology("SDK Demo", ontology_builder.asdict())
media_type will become a required field
Starting March 31, all new ontologies created with the SDK will require you to pass in a new parameter to define the media type you intend you use with the ontology (ex. Image, Pdf, etc). This will need match with the media type of the labeling project when attaching ontologies.
Create ontology from existing features
## Search feature by name in your org
regulatory_sign_feature_schema = next(client.get_feature_schemas("regulatory-sign"))
classification_feature = next(client.get_feature_schemas("Quality Issues"))
## Get feature by feature schema ID. You can get this from the UI
regulatory_sign_feature_schema = client.get_feature_schema("FEATURE_SCHEMA_ID")
print(regulatory_sign_feature_schema)
print(classification_feature)
ontology = client.create_ontology_from_feature_schemas("Ontology of shared features", [regulatory_sign_feature_schema.uid, classification_feature.uid])
Read features and ontology
It is best to reuse features and ontologies whenever you can to avoid cluttering your workspace.
#### Fetch by ID
feature_schema = client.get_feature_schema(feature_schema_cat.uid)
ontology = client.get_ontology(ontology.uid)
print(feature_schema)
print(ontology)
### Attach existing ontology with a project
editor = next(client.get_labeling_frontends(where = LabelingFrontend.name == 'editor'))
project.setup(editor, ontology.normalized)
#### Search by name
feature_schema = next(client.get_feature_schemas("cat"))
ontology = next(client.get_ontologies(ontology_name))
print(feature_schema)
print(ontology)
Complete tutorial in Python SDK
| Python Tutorial | Github | Google Colab |
|---|---|---|
| Working with features and ontologies |  |  |
| Importing Data Row metadata | | |
Updated about 2 months ago