A simple implementation of the k-nearest neighbours algorithm in Python.
Works with CSV files that are formatted as such:
- have the classifications in the first column
- columns after the first one will be considered for attributes in that order (explained further in the example below)
python SimpleKNN.py -d <data_url> -k <k_value> -q <query_features>
If a k-value is not provided, 11 is used by default. The query should be a list of values separated by a comma and no spaces.
The query features will be assumed to be in the same order as the data in the provided csv.
e.g.
python SimpleKNN.py -d weight-height.csv -k 5 --features 70,170
I've included some other sets you can use in the folder "DemoDataSets", although they are not all actually good use cases for this algorithm.
weight-height.csv : https://www.kaggle.com/mustafaali96/weight-height/version/1 heart.csv : https://www.kaggle.com/ronitf/heart-disease-uci london / seattle data : http://insideairbnb.com/get-the-data.html (I have edited the data) k-nearest neighbors algorithm : https://en.wikipedia.org/wiki/K-nearest_neighbors_algorithm