Used to dynamically load data-snack
Entity
objects from json schema.
Especially useful for complex dataset with many fields.
Data Snack can be easily installed using pypi repository.
pip install data_snack_dynamic_entity
First you need to define a dictionary containing the template - a configuration for your Entity type. Template should contain:
- the name of the new Entity type and its properties
- each property should have:
-
type
- type of the field -
default
- default value (optional property) -
optional
- true if the field is optional (optional property)
-
Notice that your template can be saved in any file of your choosing. Just make sure it's later parsed to a dictionary in the right format.
The template is defined using following schema: src/data_snack_dynamic_entity/entityTemplates.schema.json
.
{
"Car": {
"properties": {
"name": {
"type": "str"
},
"cost": {
"type": "float",
"default": 10.0,
"optional": True
}
}
}
}
Now you are ready to create entities based on your template:
from data_snack_dynamic_entity.factory import load_entities
entities = load_entities(templates=your_config)
After that entities
will contain a dictionary in the following format:
{
"entity_name": EntityType
}
Plugin was created by the Data Science team from Webinterpret.