pipeline_config

exception virtualitics_sdk.nlp.pipeline_config.NoDatasetIdxException

Bases: Exception

class virtualitics_sdk.nlp.pipeline_config.PipelineConfigurationStep(data_upload_step, show_date_filtering_selection=False, show_pipeline_dropdown=False, show_thresold_exp_drop=False)

Bases: Step

available_tasks_drop_title = 'Configure Pipeline'

classmethod create_date_filtering_components(pandas_df)

classmethod create_feature_selection_card(pandas_df)

classmethod create_pipeline_config_drop(default_model)

Generate the pipeline task dropdown based on a selected model Args:

default_model: The spacy model to use to fetch the available spacy tasks

Returns: A dropdown component that can be added to the page

default_pipeline_tasks = ['Event Extraction', 'Entity Extraction', 'Corpus Statistics']

dt_dropdown_title = '[Optional] Select Datetime Column'

dt_format_input = 'Datetime Format'

classmethod get_idx(store_interface)

get_pipe_config_or_default(store_interface)

In advanced mode, user can configure the spacy pipeline. This method can be called from other steps to get the selected value from the user or the default one. Args:

store_interface:

Returns: user input or default list of tasks

identifier_dropdown_title = 'Select Unique ID'

kpi_dropdown_title = '[Optional] Select KPI'

narrative_dropdown_title = 'Select Text Column'

run(flow_metadata)

step_title = 'Pipeline Configuration'