Models#

A collection of classes for building a SimAI model. Launching a build requires a configuration (ModelConfiguration) which defines the model properties, such as its inputs and outputs, the Global Coefficients and the Domain of Analysis, and its project. The ModelConfiguration object is, then, parsed to models.build() for launching a build.

Directory#

class ModelDirectory#

Provides a collection of methods related to building models.

build(configuration: ModelConfiguration, dismiss_data_with_fields_discrepancies: bool = False, dismiss_data_with_volume_overflow: bool = False)#

Launches a build given a configuration.

Parameters:
  • configuration (ModelConfiguration) – a ModelConfiguration object that contains the properties to be used in the build

  • dismiss_data_with_fields_discrepancies (bool) – set to True for omitting data with missing properties

  • dismiss_data_with_volume_overflow (bool) – set to True for omitting data outside the Domain of Analysis

Example

Use a previous configuration for a new build in the same project

a_project = simai.projects.get("project_A")

build_conf = a_project.last_model_configuration

new_model = simai.models.build(build_conf)

Use a previous configuration for a new build in another project

a_project = simai.projects.get("project_A")

build_conf = a_project.last_model_configuration

b_project = simai.projects.get("project_B")

# set the id of b_project as the project_id of the configuration
build_conf.project = b_project

new_model = simai.models.build(build_conf)

Model#

class Model#

Training model representation.

property configuration: ModelConfiguration#

Build configuration of a model.

property project_id: str#

The ID of the project where the model exists.