def test_train():
model = SVM()
data_filepath = os.path.dirname(__file__) + "/data/test-data.tsv"
labels_filepath = os.path.dirname(__file__) + "/data/test-data-label.tsv"
data = model.load_tsv_file(data_filepath)
labels = model.load_tsv_file(labels_filepath)
clf = model.train(data, labels)
def test_load_delimited_file():
model = SVM()
filepath = os.path.dirname(__file__) + "/data/test-data.tsv"
data = model.load_delimited_file(filepath, "\t")
expected = [[0, 9, 0], [2, 8, 0], [2, 3, 2], [0, 9, 1], [2, 1, 7],
[1, 1, 7], [0, 3, 6], [4, 1, 5], [2, 1, 7], [0, 2, 1]]
assert_equal(data.tolist(), expected)
def test_dump_model():
model = SVM()
data_filepath = os.path.dirname(__file__) + "/data/test-data.tsv"
labels_filepath = os.path.dirname(__file__) + "/data/test-data-label.tsv"
data = model.load_tsv_file(data_filepath)
labels = model.load_tsv_file(labels_filepath)
clf = model.train(data, labels)
pickled_filepath = os.path.dirname(__file__) + "/data/trained-svm.pickle"
pickled_string = model.dump_model(pickled_filepath)
assert (isinstance(pickled_filepath, str))
assert_equal(len(pickled_string), 1734)
#!/usr/bin/env python3
#coding:utf-8
import os
import sys
PROJECT_HOME = os.path.dirname(os.path.abspath(__file__)) + "/../"
sys.path.append(PROJECT_HOME)
from webapp.recommend.svm import SVM
model = SVM()
## load test data
data_filepath = os.path.dirname(
__file__) + "/../../resources/data/test-data.tsv"
labels_filepath = os.path.dirname(
__file__) + "/../../resources/data/test-label.tsv"
data = model.load_tsv_file(data_filepath)
labels = model.load_tsv_file(labels_filepath)
## restore model from file
result = model.predict_with_default_dumped_model(data)
num_true = 0
num_false = 0
for i in range(0, len(labels)):
if labels[i] == result[i]:
num_true += 1
else:
num_false += 1
def test_predict_with_default_dumped_model():
model = SVM()
target = [0, 2, 1]
result = model.predict_with_default_dumped_model(target)
assert_equal(result, [3])
def test_predict_with_dumped_model():
model = SVM()
target = [0, 2, 1]
filepath = os.path.dirname(__file__) + "/data/trained-svm.pickle"
result = model.predict_with_dumped_model(target, filepath)
assert_equal(result, [2])
def test_loads_model_by_pickle():
pickled_filepath = os.path.dirname(__file__) + "/data/trained-svm.pickle"
model = SVM()
clf = model.load_model_by_pickle(pickled_filepath)
predict = clf.predict([0, 9, 0])
assert_equal(predict, [3])