Metadata-Version: 2.1
Name: ucimlr
Version: 0.0.2
Summary: Easy access to datasets from the UCI Machine Learning Repository
Home-page: https://github.com/isacarnekvist/ucimlr
Author: Isac Arnekvist
Author-email: isac.arnekvist@gmail.com
License: UNKNOWN
Platform: UNKNOWN
Requires-Python: >=3.6
Description-Content-Type: text/markdown
Requires-Dist: numpy (>=1.18.2)
Requires-Dist: pandas (>=1.0.3)
Requires-Dist: sklearn (>=0.0)
Requires-Dist: unlzw (>=0.1.1)
Requires-Dist: xlrd (>=1.0.0)

# UCI Machine Learning Repository - API
This package provides easy access to a number of datasets from
the UCI Machine Learning repository.

More documentation will come.

## Citation
Make sure you visit UCI and the specific datasets you use to
make sure that they are cited properly.

## Basic usage
```
import ucimlr

dataset = ucimlr.datasets.Abalone('datasets', download=True)
# Datasets are either 'classification' or 'regression':
print(dataset.type_)
>>> regression

# The attribute 'num_features' show the number of dimensions
# of the independent variable.
print(dataset.num_features)
>>> 10
print(dataset.x.shape)
>>> (3341, 10)

# For datasets of type 'regression', you can inspect the
# dimensionality of the output:
print(dataset.num_targets)
>>> 1
print(dataset.y.shape)
>>> (3341, 1)

# For a classification dataset, the only difference is
# the shape of the dependent variable
dataset = ucimlr.datasets.CardDefault('datasets', download=True)
print(dataset.type_)
>>> classification
# Query number of classes instead of number of features
print(dataset.num_classes)
>>> 2
print(dataset.y.shape)
>>> (24000,)
```

