gokinjo 0.1.0

k-NN feature extraction utility

# gokinjo

• A feature extraction library based on k-nearest neighbor algorithm in Python - k-NN based feature has experience of being used in 1st place solution of Kaggle competition

• Switchable backend of k-NN algorithm - [scikit-learn](https://github.com/scikit-learn/scikit-learn) (default) - [annoy](https://github.com/spotify/annoy)

• FYI: ‘gokinjo’ is meant neighborhood in japanese.

### Prerequisite

• Python 3.6 or later

• setuptools >= 30.0.3.0

### How to install

#### From PyPI

`bash $ pip install gokinjo `

##### With annoy backend

`bash $ pip install "gokinjo[annoy]" `

#### From source code

`bash $ pip install git+https://github.com/momijiame/gokinjo.git `

### Usage example

• Please see [examples](https://github.com/momijiame/gokinjo/tree/master/examples) in GitHub repository.

### How to setup of development environment

`bash $ pip install -e ".[develop]" $ pytest `

### References

• The competition which k-NN feature was used on 1st place solution - https://www.kaggle.com/c/otto-group-product-classification-challenge/discussion/14335

• R implementation - https://github.com/davpinto/fastknn/blob/master/R/extract.R

• Another Python implementation - https://github.com/upura/knnFeat