# recs.export_csv(os.path.join(path, "new_csv_file.csv"),delimiter=",",line_terminator='\n', header=True)
# for accessing AWS keys in .env file
dotenv_path = join(dirname(__file__),'.env')
load_dotenv(dotenv_path)
access_key = os.getenv("AWS_ACCESS_KEY_ID")
secret_key = os.getenv("AWS_SECRET_ACCESS_KEY")
# # Read S3 bucket file content
conn = S3Connection(access_key, secret_key)
bucket = conn.get_bucket('niche-travel')
user_recs = bucket.get_key('user_recs.csv')
user_recs = user_recs.generate_url(120, method="GET")
data = gl.SFrame.read_csv(user_recs)
print data
m = gl.recommender.item_similarity_recommender.create(data, user_id='user_id', item_id='place_id', target="response", user_data=None, item_data=None, nearest_items=None, similarity_type='pearson', training_method='auto', threshold=0.001, only_top_k=100, random_seed=0, verbose=True)
recs = m.recommend()
# Write to different S3 bucket
k = Key(bucket)
k.new_key = 'new_user_recs.csv'
k = k.generate_url(30, method="POST")
recs.save("s3://niche-travel/new_user_recs.csv", format="csv")
