Ignore:
Timestamp:
Jul 30, 2019, 8:28:06 PM (5 years ago)
Author:
Murray Heymann <heymann.murray@…>
Branches:
(u'spielwiese', 'fe61d9c35bf7c61f2b6cbf1b56e25e2f08d536cc')
Children:
fece1392f8e9ff07b64d0ed4e5ec57bfa6dbf258
Parents:
da892581a52069935f084604d05c0ecd6d19d5c9
Message:
Finish migrating to unittest framework
File:
1 edited

Legend:

Unmodified
Added
Removed

  • machine_learning/model/predictor.py

    rda8925 r50e872  
    6666            ret_list.append(file)
    6767        return np.array(ret_list)
    68 
    69 
    70 def basic_vector_tests():
    71     """
    72     Some basic sanity tests
    73     """
    74     predictor = HelpPagePredictor()
    75     vector1 = normalise_vector(np.array([1, 4, 10]))
    76     vector2 = normalise_vector(np.array([2, 3, 1]))
    77     vector3 = normalise_vector(np.array([3, 9, 3]))
    78 
    79     vectors = np.array([vector1, vector2, vector3])
    80     files = np.array(["file1", "file2", "file3"])
    81     print(vectors)
    82     print(files)
    83     print()
    84 
    85     testvec = normalise_vector(np.array([1, 1, 1]))
    86     print("test vector:")
    87     print(testvec)
    88     print()
    89 
    90     print("distance to 1")
    91     print(vector_distance(testvec, vector1))
    92     print()
    93     print("distance to 2")
    94     print(vector_distance(testvec, vector2))
    95     print()
    96     print("distance to 3")
    97     print(vector_distance(testvec, vector3))
    98     print()
    99 
    100     predictor.fit(vectors, files)
    101     prediction = predictor.predict(np.array([testvec]))
    102     print("Prediction:")
    103     print(prediction)
    104     print()
    105 
    106 
    107 def main():
    108     """
    109     Run some basic tests
    110     """
    111     print("Running some tests")
    112 
    113     basic_vector_tests()
    114 
    115     dictionary = read_dictionary(KEYWORDS_FILE)
    116 
    117     start = time.time()
    118     vectors, file_list = create_table(dictionary=dictionary)
    119     end = time.time()
    120     print(end - start, "seconds to create_table")
    121 
    122     predictor = HelpPagePredictor()
    123     predictor.fit(vectors, file_list)
    124 
    125     start = time.time()
    126     test_vec = count_occurances("extract.lib", dictionary)
    127     prediction = predictor.predict(np.array([test_vec]))
    128     end = time.time()
    129     print(end - start, "seconds to make prediction")
    130     print(prediction)
    131     print()
    132 
    133     print("prediction for zero vector")
    134     start = time.time()
    135     zerovec = np.zeros(len(dictionary) - 2)
    136     prediction = predictor.predict(np.array([zerovec]))
    137     end = time.time()
    138     print(end - start, "seconds to make prediction")
    139     print(prediction)
    140     print()
    141 
    142     if len(sys.argv) >= 2:
    143         for i in range(len(sys.argv)):
    144             if i == 0:
    145                 continue
    146             if not os.path.isfile(sys.argv[i]):
    147                 continue
    148             print("predicting for file", sys.argv[i])
    149             start = time.time()
    150             test_vec = count_occurances(sys.argv[i], dictionary)
    151             prediction = predictor.predict(np.array([test_vec]))
    152             end = time.time()
    153             print(end - start, "seconds to make prediction")
    154             print(prediction)
    155             print()
    156 
    157 
    158 if __name__ == '__main__':
    159     #cProfile.run("main()")
    160     main()
Note: See TracChangeset for help on using the changeset viewer.