def test_input_diff_constraint(self):
# API and Cython object test. Replicates command line:
# ./MPP -file input.txt -thr 0.001 -att input_att1.txt -lg 20 -ug 1000 -ls 800 - us 3700 -att input_att2.txt -la 20 -ua 80 -lm 30 -um 70 -out -write BMS_patt.txt
# input on Main.cpp and verifies output with data captured from original implementation
# Similar to default but all lower constraints lowered by 10 all upper constraints raised 10.
# Significantly different results to default.
# Seq2Pat
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
seq2pat = Seq2Pat(sequences)
# Load Attributes
attribute_file = self.DATA_DIR + "input_att1.txt"
attr1_data = read_data(attribute_file)
att1 = Attribute(attr1_data)
attribute_file = self.DATA_DIR + "input_att2.txt"
attr2_data = read_data(attribute_file)
att2 = Attribute(attr2_data)
cts1 = seq2pat.add_constraint(20 <= att1.gap() <= 1000)
cts2 = seq2pat.add_constraint(800 <= att1.span() <= 3700)
cts3 = seq2pat.add_constraint(20 <= att2.average() <= 80)
cts4 = seq2pat.add_constraint(30 <= att2.median() <= 70)
test_patterns = seq2pat.get_patterns(.001)
results_file = self.DATA_DIR + "diff_constraints_results.txt"
control_patterns = read_data(results_file)
sorted_control = sort_pattern(control_patterns)
self.assertListEqual(sorted_control, test_patterns)
def test_input_no_upper_constraint(self):
# API and Cython object test. Replicates command line:
# ./MPP -file input.txt -thr 0.01 -att input_att1.txt -lg 30 -ls 900 -att input_att2.txt -la 30 -lm 40 -out -write BMS_patt.txt
# input on Main.cpp and verifies output with data captured from original implementation
# Similar to default but no upper constraints imposed
# Seq2Pat
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
seq2pat = Seq2Pat(sequences)
# Load Attributes
attribute_file = self.DATA_DIR + "input_att1.txt"
attr1_data = read_data(attribute_file)
att1 = Attribute(attr1_data)
attribute_file = self.DATA_DIR + "input_att2.txt"
attr2_data = read_data(attribute_file)
att2 = Attribute(attr2_data)
cts1 = seq2pat.add_constraint(30 <= att1.gap())
cts2 = seq2pat.add_constraint(900 <= att1.span())
cts3 = seq2pat.add_constraint(30 <= att2.average())
cts4 = seq2pat.add_constraint(40 <= att2.median())
test_patterns = seq2pat.get_patterns(.01)
results_file = self.DATA_DIR + "no_upper_constraint_results.txt"
control_patterns = read_data(results_file)
sorted_control = sort_pattern(control_patterns)
self.assertListEqual(sorted_control, test_patterns)
def test_input_one_constraint(self):
# API and Cython object test. Replicates command line:
# ./MPP -file input.txt -thr 0.001 -att input_att1.txt -lg 30 -ug 900 -ls 900 - us 3600 -out -write BMS_patt.txt
# input on Main.cpp and verifies output with data captured from original implementation
# Similar to default but with constraints on a single attribute- input_att1.txt
# Seq2Pat
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
seq2pat = Seq2Pat(sequences)
# Load Attributes
attribute_file = self.DATA_DIR + "input_att1.txt"
attr1_data = read_data(attribute_file)
att1 = Attribute(attr1_data)
cts1 = seq2pat.add_constraint(30 <= att1.gap() <= 900)
cts2 = seq2pat.add_constraint(900 <= att1.span())
test_patterns = seq2pat.get_patterns(.001)
results_file = self.DATA_DIR + "one_constraint_results.txt"
control_patterns = read_data(results_file)
sorted_controls = sort_pattern(control_patterns)
self.assertListEqual(sorted_controls, test_patterns)
self.assertFalse(test_patterns == read_data(self.DATA_DIR +
"default_results.txt"))
def test_usage(self):
# Pattern data
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
# print("Patterns: ", sequences[:5])
# Attribute data
attribute_file = self.DATA_DIR + "input_att1.txt"
attribute_1 = read_data(attribute_file)
# print("Attribute_1: ", attribute_1[:5])
# Sequential pattern finder
seq2pat = Seq2Pat(sequences)
# Constraints on attribute 1
att1 = Attribute(attribute_1)
avg_constraint = seq2pat.add_constraint(5 <= att1.average())
gap_constraint = seq2pat.add_constraint(att1.gap() <= 10)
median_constraint = seq2pat.add_constraint(10 <= att1.median() <= 15)
span_constraint = seq2pat.add_constraint(att1.span() <= 20)
# Print constraint store
# seq2pat.__str__()
seq2pat.get_patterns(min_frequency=100)
def test_input(self):
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
self.assertTrue(type(sequences) == list)
self.assertTrue(len(sequences) == 52619)
self.assertTrue(len(sequences[0]) == 5)
# print(sequences[:5])
attribute_file = self.DATA_DIR + "input_att1.txt"
attribute_1 = read_data(attribute_file)
self.assertTrue(type(attribute_1) == list)
self.assertTrue(len(attribute_1) == 52619)
self.assertTrue(len(attribute_1[0]) == 5)
def test_input_no_constraint(self):
# API and Cython object test. Replicates command line:
# ./MPP -file input.txt -thr 0.01 -out
# input on Main.cpp and verifies output with data captured from original implementation
# Unconstrained call. Significantly different and larger results.
# Seq2Pat
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
seq2pat = Seq2Pat(sequences)
test_patterns = seq2pat.get_patterns(.01)
results_file = self.DATA_DIR + "no_constraints_results.txt"
control_patterns = read_data(results_file)
sorted_results = sort_pattern(control_patterns)
self.assertListEqual(sorted_results, test_patterns)
self.assertFalse(test_patterns == read_data(self.DATA_DIR +
"default_results.txt"))
def test_input_item_variables(self):
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
m = get_max_column_size(sequences)
n = len(sequences)
l = get_max_value(sequences)
self.assertEqual(161, m)
self.assertEqual(52619, n)
self.assertEqual(3340, l)
def test_seq2patfinder_default(self):
# API and Cython object test.
# Replicates command line:
# > ./MPP.exe
# -file input.txt
# -thr 0.001
# -att input_att1.txt -lg 30 -ug 900 -ls 900 - us 3600
# -att input_att2.txt -la 30 -ua 70 -lm 40 -um 60
# -out -write BMS_patt.txt
# input on Main.cpp and verifies output with data captured from original implementation
# Seq2Pat
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
seq2pat = Seq2Pat(sequences)
# Load Attributes
attribute_file = self.DATA_DIR + "input_att1.txt"
attr1_data = read_data(attribute_file)
att1 = Attribute(attr1_data)
attribute_file = self.DATA_DIR + "input_att2.txt"
attr2_data = read_data(attribute_file)
att2 = Attribute(attr2_data)
cts1 = seq2pat.add_constraint(30 <= att1.gap() <= 900)
cts2 = seq2pat.add_constraint(900 <= att1.span())
cts3 = seq2pat.add_constraint(30 <= att2.average() <= 70)
cts4 = seq2pat.add_constraint(40 <= att2.median() <= 60)
test_pf = seq2pat._get_cython_imp(-1)
self.assertListEqual([30], test_pf.lgap)
self.assertListEqual([900], test_pf.ugap)
self.assertListEqual([30], test_pf.lavr)
self.assertListEqual([70], test_pf.uavr)
self.assertListEqual([900], test_pf.lspn)
self.assertListEqual([], test_pf.uspn)
self.assertListEqual([40], test_pf.lmed)
self.assertListEqual([0], test_pf.ugapi)
self.assertListEqual([0], test_pf.lgapi)
self.assertListEqual([], test_pf.uspni)
self.assertListEqual([0], test_pf.lspni)
self.assertListEqual([1], test_pf.uavri)
self.assertListEqual([1], test_pf.lavri)
self.assertListEqual([1], test_pf.umedi)
self.assertListEqual([1], test_pf.lmedi)
self.assertListEqual([2, 0], test_pf.num_minmax)
self.assertListEqual([0, 2], test_pf.num_avr)
self.assertListEqual([0, 2], test_pf.num_med)
self.assertListEqual([0], test_pf.tot_gap)
self.assertListEqual([0], test_pf.tot_spn)
self.assertListEqual([1], test_pf.tot_avr)
self.assertEqual(161, test_pf.M)
self.assertEqual(52619, test_pf.N)
self.assertEqual(3340, test_pf.L)
self.assertListEqual([284871, 100], test_pf.max_attrs)
self.assertListEqual([0, 1], test_pf.min_attrs)
test_patterns = seq2pat.get_patterns(.001)
# Consistency sanity check
dup_patterns = seq2pat.get_patterns(.001)
self.assertListEqual(test_patterns, dup_patterns)
results_file = self.DATA_DIR + "default_results.txt"
control_patterns = read_data(results_file)
sorted_control = sort_pattern(control_patterns)
self.assertListEqual(sorted_control, test_patterns)
# Remove constraint and test
cts5 = seq2pat.remove_constraint(40 <= att2.median() <= 60)
ct6 = seq2pat.remove_constraint(30 <= att2.average() <= 70)
test_pf = seq2pat._get_cython_imp(-1)
self.assertListEqual([], test_pf.umedi)
self.assertListEqual([], test_pf.lmedi)
self.assertListEqual([], test_pf.uavr)
self.assertListEqual([], test_pf.uavri)
self.assertListEqual([0], test_pf.num_med)
one_constraint_result = seq2pat.get_patterns(.001)
results_file = self.DATA_DIR + "one_constraint_results.txt"
control_patterns = read_data(results_file)
sorted_controls = sort_pattern(control_patterns)
self.assertListEqual(sorted_controls, one_constraint_result)
def test_setter(self):
# Testing cython object setters and getters
python_seq2pat = stp.PySeq2pat()
patterns_file = self.DATA_DIR + "input.txt"
sequences = read_data(patterns_file)
seq2pat = Seq2Pat(sequences)
python_seq2pat.lgap = [30]
python_seq2pat.ugap = [900]
python_seq2pat.lspn = [77]
python_seq2pat.uspn = [9, 80]
python_seq2pat.lavr = [9, 88]
python_seq2pat.uavr = [7]
python_seq2pat.lmed = [9, 9, 8]
python_seq2pat.umed = [99999]
self.assertListEqual(python_seq2pat.lgap, [30])
self.assertListEqual(python_seq2pat.ugap, [900])
self.assertListEqual(python_seq2pat.lspn, [77])
self.assertListEqual(python_seq2pat.uspn, [9, 80])
self.assertListEqual(python_seq2pat.lavr, [9, 88])
self.assertListEqual(python_seq2pat.uavr, [7])
self.assertListEqual(python_seq2pat.lmed, [9, 9, 8])
self.assertListEqual(python_seq2pat.umed, [99999])
python_seq2pat.lgapi = [0]
python_seq2pat.ugapi = [0]
python_seq2pat.lspni = [1]
python_seq2pat.uspni = [1, 0]
python_seq2pat.lavri = [0, 1]
python_seq2pat.uavri = [0]
python_seq2pat.lmedi = [0, 1, 2]
python_seq2pat.umedi = [2]
#
self.assertListEqual(python_seq2pat.lgapi, [0])
self.assertListEqual(python_seq2pat.ugapi, [0])
self.assertListEqual(python_seq2pat.lspni, [1])
self.assertListEqual(python_seq2pat.uspni, [1, 0])
self.assertListEqual(python_seq2pat.lavri, [0, 1])
self.assertListEqual(python_seq2pat.uavri, [0])
self.assertListEqual(python_seq2pat.lmedi, [0, 1, 2])
self.assertListEqual(python_seq2pat.umedi, [2])
python_seq2pat.num_minmax = [0, 0, 0]
python_seq2pat.num_avr = [1, 1, 1]
python_seq2pat.num_med = [0, 1, 2]
python_seq2pat.tot_gap = [0, 1, 0]
python_seq2pat.tot_spn = [2, 2, 2]
python_seq2pat.tot_avr = [0, 1, 1]
self.assertListEqual(python_seq2pat.num_minmax, [0, 0, 0])
self.assertListEqual(python_seq2pat.num_avr, [1, 1, 1])
self.assertListEqual(python_seq2pat.num_med, [0, 1, 2])
self.assertListEqual(python_seq2pat.tot_gap, [0, 1, 0])
self.assertListEqual(python_seq2pat.tot_spn, [2, 2, 2])
self.assertListEqual(python_seq2pat.tot_avr, [0, 1, 1])
python_seq2pat.num_att = 3
python_seq2pat.N = 200
python_seq2pat.M = 999
python_seq2pat.L = 89
python_seq2pat.theta = 89
self.assertEqual(python_seq2pat.num_att, 3)
self.assertEqual(python_seq2pat.N, 200)
self.assertEqual(python_seq2pat.M, 999)
self.assertEqual(python_seq2pat.L, 89)
self.assertEqual(python_seq2pat.theta, 89)