Fetching and Loading Data¶

This is a short tutorial showing how to fetch and load data in context.

In [ ]:

!pip install relational-datasets

!pip install relational-datasets

fetch()¶

fetch() always returns the path to a specific version of a dataset.

If the dataset is not available in the cache, it is downloaded from the srlearn/datasets GitHub repository.

In [1]:

from relational_datasets import fetch

fetch('toy_cancer', 'v0.0.5')

from relational_datasets import fetch fetch('toy_cancer', 'v0.0.5')

Out[1]:

'/Users/hayesall/relational_datasets/toy_cancer_v0.0.5.zip'

load()¶

This function uses fetch() internally, so if a dataset is unavailable it is downloaded.

load() returns train and test folds in the form two namedtuples with pos, neg, and facts keys:

RelationalDataset = namedtuple("RelationalDataset", ['pos', 'neg', 'facts'])

In [2]:

from relational_datasets import load

train, test = load('toy_cancer', 'v0.0.5')

train.pos

from relational_datasets import load train, test = load('toy_cancer', 'v0.0.5') train.pos

Out[2]:

['cancer(alice).', 'cancer(bob).', 'cancer(chuck).', 'cancer(fred).']

---

Last update: June 20, 2022