kedro.extras.datasets.svmlight.SVMLightDataSet

SVMLightDataSet loads/saves data from/to a svmlight/libsvm file using an underlying filesystem (e.g.: local, S3, GCS). It uses sklearn functions dump_svmlight_file to save and load_svmlight_file to load a file.

Data is loaded as a tuple of features and labels. Labels is NumPy array, and features is Compressed Sparse Row matrix.

This format is a text-based format, with one sample per line. It does not store zero valued features hence it is suitable for sparse datasets.

This format is used as the default format for both svmlight and the libsvm command line programs.

Example usage for the YAML API:

svm_dataset:
  type: svmlight.SVMLightDataSet
  filepath: data/01_raw/location.svm
  load_args:
    zero_based: False
  save_args:
    zero_based: False

cars:
  type: svmlight.SVMLightDataSet
  filepath: gcs://your_bucket/cars.svm
  fs_args:
    project: my-project
  credentials: my_gcp_credentials
  load_args:
    zero_based: False
  save_args:
    zero_based: False

Example usage for the Python API:

>>> from kedro.extras.datasets.svmlight import SVMLightDataSet
>>> import numpy as np
>>>
>>> # Features and labels.
>>> data = (np.array([[0, 1], [2, 3.14159]]), np.array([7, 3]))
>>>
>>> data_set = SVMLightDataSet(filepath="test.svm")
>>> data_set.save(data)
>>> reloaded_features, reloaded_labels = data_set.load()
>>> assert (data[0] == reloaded_features).all()
>>> assert (data[1] == reloaded_labels).all()