itemRecommendations.py

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import sklearn.metrics as metrics
import numpy as np
from sklearn.neighbors import NearestNeighbors


 
# This contains a function `find_similar_items` which calculates the most similar items using cosine similarity.
# I figured this made more sense than user based recommendations because someone who hasn't bought any items before would have
# zero similarity to other users, right?
# 
# I generally followed a tutorial here: https://github.com/csaluja/JupyterNotebooks-Medium/blob/master/CF%20Recommendation%20System-Examples.ipynb



# user/product array - 1 for purchase, 0 for no purchase history
# presumably could have another service that constructs this and then load it from a csv?
M = np.asarray([[1, 0, 0, 1, 1, 1, 1, 0],
                [0, 1, 0, 0, 1, 0, 1, 0],
                [0, 1, 0, 0, 1, 1, 1, 1],
                [0, 0, 0, 1, 1, 1, 1, 0],
                [1, 0, 1, 1, 1, 1, 0, 1],
                [1, 0, 0, 1, 1, 1, 0, 0]])
M = pd.DataFrame(M)

def find_similar_items(item_id, ratings, k=3):
    similarities = []
    indices = []
    ratings = ratings.T
    model_knn = NearestNeighbors(metric='cosine', algorithm='brute')
    model_knn.fit(ratings)

    distances, indices = model_knn.kneighbors(
        ratings.iloc[item_id-1, :].values.reshape(1, -1), n_neighbors=k+1)
    similarities = 1-distances.flatten()
    print('{0} most similar items for item {1}:\n'.format(k, item_id))
    for i in range(0, len(indices.flatten())):
        if indices.flatten()[i]+1 == item_id:
            continue

        else:
            print('{0}: Item {1} :, with similarity of {2}'.format(i, indices.flatten()[i]+1, similarities.flatten()[i]))

    return similarities, indices

similarities, indices = find_similar_items(3, M)