search

ITMO-NSS-team / fedot-examples

The repository contains examples for the AutoML framework FEDOT
https://github.com/aimclub/FEDOT
47 stars 14 forks source link

No data warning #6

Open rnjv opened 2 years ago

rnjv commented 2 years ago

I was trying to follow the Jupyter example and when I try to do it on MultiModalData and my own dataset I see the following:

Pipeline assessment warning: No data for primary node sparse_lagged. Continue. Pipeline assessment warning: No data for primary node ar. Continue. Pipeline assessment warning: No data for primary node ar. Continue. Pipeline assessment warning: No data for primary node sparse_lagged. Continue

Any advice how to format the dataset especially to trigger the ar node?

nicl-nno commented 2 years ago

Hi!

Please provide the example of your code and data to reproduce this error. It looks like the MultiModalData is formatted improperly in your case.

rnjv commented 2 years ago

HI!

MultiModalData configured as here

I used the following to trigger the composer:

mutation_types = [parameter_change_mutation, MutationTypesEnum.simple, MutationTypesEnum.reduce]
optimiser_parameters = GPGraphOptimiserParameters(mutation_types=mutation_types)
#%%
init_pipeline = get_two_branches_pipeline(lagged_param_1=96, lagged_param_2=4)
#%%
metric_function = MetricsRepository().metric_by_id(RegressionMetricsEnum.MAE)
builder = GPComposerBuilder(task=task). \
    with_optimiser_parameters(optimiser_parameters). \
    with_requirements(composer_requirements). \
    with_metrics(metric_function).with_initial_pipeline(init_pipeline)
composer = builder.build()

obtained_pipeline = composer.compose_pipeline(data=train_dataset, is_visualise=False)