jubarecommender example (programming language)

analyze.py

# -*- coding: utf-8 -*-

from jubatus.recommender import client
from jubatus.recommender import types
from jubatus.common import Datum
import sys


def print_result(key, score_list):
    print('{} is similar to:'.format(key))
    for i, row in enumerate(score_list):
        print('{:>3} {:<10} ({})'.format(i+1, row.id, row.score))


if __name__ == '__main__':

    param = sys.argv
    if len(param) < 3:
        quit()

    recommender = client.Recommender('127.0.0.1', 9199, 'sample')

    result = recommender.similar_row_from_id(str(param[1]), int(param[2]))
    print_result(param[1], result)

config.json

{
  "method": "inverted_index",
  "converter": {
    "string_filter_types": {},
    "string_filter_rules": [],
    "num_filter_types": {},
    "num_filter_rules": [],
    "string_types": {},
    "string_rules": [
      {"key": "*", "type": "space", "sample_weight": "bin", "global_weight": "bin"}
    ],
    "num_types": {},
    "num_rules": []
  },
  "parameter": {}
}

programming_languages.tsv

update.py

# -*- coding: utf-8 -*-

from jubatus.recommender import client
from jubatus.recommender import types
from jubatus.common import Datum


def read_tsv(data_path):
    with open(data_path) as f:
        for line in f:
            yield line.split('\t')


if __name__ == '__main__':
    recommender = client.Recommender('127.0.0.1', 9199, 'sample')

    it = read_tsv('./programming_languages.tsv')
    next(it) # skip header
    for row in it:
        data_dict = {}
        language = row[0]
        data_dict['Intentd use']       = row[1]
        data_dict['Imperative']        = row[2]
        data_dict['Object-oriented']   = row[3]
        data_dict['Functional']        = row[4]
        data_dict['Procedural']        = row[5]
        data_dict['Generic']           = row[6]
        data_dict['Reflective']        = row[7]
        data_dict['Event-driven']      = row[8]
        data_dict['Other paradigm(s)'] = row[9]

        d = Datum(data_dict)
        recommender.update_row(language, d)