README.md

metadata

language:
  - en
license: mit
size_categories:
  - 1K<n<10K
task_categories:
  - time-series-forecasting
pretty_name: Next purchase day
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: validation
        path: data/validation-*
dataset_info:
  features:
    - name: start
      dtype: timestamp[us]
    - name: target
      sequence: float32
    - name: item_id
      dtype: string
    - name: feat_static_cat
      sequence: uint64
    - name: feat_dynamic_real
      sequence:
        sequence: float32
  splits:
    - name: train
      num_bytes: 11055
      num_examples: 34
    - name: test
      num_bytes: 12143
      num_examples: 34
    - name: validation
      num_bytes: 11599
      num_examples: 34
  download_size: 29957
  dataset_size: 34797