components/dataprep/data_split.yaml

$schema: https://azuremlschemas.azureedge.net/latest/commandComponent.schema.json

name: data_split_alzheimers
version: 0.2.0
description: Reads multiple datasets of Alzheimer's images and splits them into training and testing datasets.
display_name: Data Splitting for Alzheimer's Dataset
is_deterministic: true
type: command

code: ./code
command: 'python traintestsplit.py
  --mild_data ${{inputs.mild_data}}
  --moderate_data ${{inputs.moderate_data}}
  --non_data ${{inputs.non_data}}
  --very_mild_data ${{inputs.very_mild_data}}
  --split_size ${{inputs.train_test_split_factor}}
  --training_data_output ${{outputs.training_data}}
  --testing_data_output ${{outputs.testing_data}}
  '

inputs:
  mild_data:
    optional: false
    type: uri_folder
    description: Path to the MildDemented dataset folder.
  moderate_data:
    optional: false
    type: uri_folder
    description: Path to the ModerateDemented dataset folder.
  non_data:
    optional: false
    type: uri_folder
    description: Path to the NonDemented dataset folder.
  very_mild_data:
    optional: false
    type: uri_folder
    description: Path to the VeryMildDemented dataset folder.
  train_test_split_factor:
    optional: false
    type: number
    description: Percentage of data to use for testing (e.g., 20 for 20% testing data).

outputs:
  testing_data:
    type: uri_folder
    description: Path to the output folder for testing data.
  training_data:
    type: uri_folder
    description: Path to the output folder for training data.

resources:
  instance_count: 1

environment: azureml:aml-preprocessing-cli:0.1.0