def test_offset_1part_5(self):
# already regular offset interval (but some missing) to larger one
in_val = [pandas.Series(['a', 'b', 'c'], index=[0.0, 0.5, 1.5], name=('Indexer', '0'))]
expected = pandas.Series(['a', 'b', 'c'], index=[0.0, 1.0, 2.0], name=('Indexer', '0'))
offset_interval = 1.0
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(1, len(actual.columns)) # same number of columns?
actual = actual['0']
self.assertSequenceEqual(list(expected.values), list(actual.values)) # same rows?
self.assertSequenceEqual(list(expected.index), list(actual.index)) # same index?
def test_offset_1part_2(self):
# input is expected output
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.5, 1.0, 1.5], name=('Indexer', '0'))]
expected = pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.5, 1.0, 1.5], name=('Indexer', '0'))
offset_interval = 0.5
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(1, len(actual.columns)) # same number of columns?
actual = actual['0']
self.assertSequenceEqual(list(expected.values), list(actual.values)) # same rows?
self.assertSequenceEqual(list(expected.index), list(actual.index)) # same index?
def test_offset_1part_1(self):
# 0 length
in_val = [pandas.Series(name=('Indexer', '0'))]
expected = pandas.Series(name=('Indexer', '0'))
offset_interval = 0.5
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(len(in_val), len(actual.columns)) # same number of columns?
actual = actual['0']
self.assertEqual(len(expected), len(actual)) # same number of rows?
self.assertEqual(list(expected.index), list(actual.index)) # same row names?
def test_offset_1part_10(self):
# targeted test for end-of-piece: when last thing doesn't land on an observed offset
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.4, 1.1, 2.1], name=('N', '0'))]
expected = pandas.Series(['a', 'b', 'b', 'c', 'c', 'd'], name=('N', '0'),
index=[0.0, 0.5, 1.0, 1.5, 2.0, 2.5])
offset_interval = 0.5
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(1, len(actual.columns)) # same number of columns?
actual = actual['0']
self.assertSequenceEqual(list(expected.values), list(actual.values)) # same rows?
self.assertSequenceEqual(list(expected.index), list(actual.index)) # same index?
def test_offset_1part_4b(self):
# already regular offset interval to a very small one
in_val = [pandas.Series(['a', 'b'], index=[0.0, 0.5], name=('Indexer', '0'))]
expected = pandas.Series(['a', 'a', 'a', 'a', 'b'], name=('Indexer', '0'),
index=[0.0, 0.125, 0.25, 0.375, 0.5])
offset_interval = 0.125
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(1, len(actual.columns)) # same number of columns?
actual = actual['0']
self.assertSequenceEqual(list(expected.values), list(actual.values)) # same rows?
self.assertSequenceEqual(list(expected.index), list(actual.index)) # same index?
def test_offset_xparts_0a(self):
# 0 length, many parts
in_val = [pandas.Series(name=('N', str(i))) for i in range(4)]
expected = pandas.DataFrame({str(i): pandas.Series() for i in range(4)})
offset_interval = 12.0
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(len(expected.columns), len(actual.columns))
expected = expected.fillna(value='None') # avoid having to use isnan()
actual = actual.fillna(value='None')
for partname in expected.columns:
self.assertSequenceEqual(list(expected[partname].values), list(actual[partname].values))
self.assertSequenceEqual(list(expected[partname].index), list(actual[partname].index))
def test_offset_xparts_1(self):
# input is expected output; 2 parts
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.5, 1.0, 1.5], name=('N', '0')),
pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.5, 1.0, 1.5], name=('N', '1'))]
expected = {'0': pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.5, 1.0, 1.5]),
'1':pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.5, 1.0, 1.5])}
expected = pandas.DataFrame(expected)
offset_interval = 0.5
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(len(expected.columns), len(actual.columns))
expected = expected.fillna(value='None') # avoid having to use isnan()
actual = actual.fillna(value='None')
for partname in expected.columns:
self.assertSequenceEqual(list(expected[partname].values), list(actual[partname].values))
self.assertSequenceEqual(list(expected[partname].index), list(actual[partname].index))
def test_run_2_a(self):
# checks if the index and number of columns are correct if one of the passed series is empty.
in_val = [pandas.Series(['A', 'B'], name=('N', '0')), pandas.Series(name=('N', '1'))]
settings = {'quarterLength': 1.0, 'method': 'ffill'}
actual = FilterByOffsetIndexer(in_val, settings).run()
self.assertListEqual(list(actual.index), [0.0, 1.0])
self.assertEqual(len(actual.columns), 2)
def test_offset_xparts_2(self):
# input is expected output; 10 parts
letters = ['a', 'b', 'c', 'd']
offsets = [0.0, 0.5, 1.0, 1.5]
in_val = [pandas.Series(letters, index=offsets, name=('N', str(_))) for _ in range(10)]
expected = {str(i): pandas.Series(letters, index=offsets) for i in range(10)}
expected = pandas.DataFrame(expected)
offset_interval = 0.5
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(len(expected.columns), len(actual.columns))
expected = expected.fillna(value='None') # avoid having to use isnan()
actual = actual.fillna(value='None')
for partname in expected.columns:
self.assertSequenceEqual(list(expected[partname].values), list(actual[partname].values))
self.assertSequenceEqual(list(expected[partname].index), list(actual[partname].index))
def test_run_1_a(self):
# try statement correctly finds minimum start_offset and whole index when no series is empty.
in_val = [pandas.Series(['A', 'B'], name=('N', '0')),
pandas.Series(['B', 'C'], index=[1, 2], name=('N', '1'))]
settings = {'quarterLength': 1.0, 'method': 'ffill'}
actual = FilterByOffsetIndexer(in_val, settings).run()
self.assertListEqual(list(actual.index), [0.0, 1.0, 2.0])
self.assertEqual(len(actual.columns), 2)
def test_run_1_b(self):
# same as test_run_1_a but with a non-zero starting point and non-overlapping indecies.
in_val = [pandas.Series(['A', 'B'], index=[4, 5], name=('N', '0')),
pandas.Series(['C', 'D'], index=[9, 10], name=('N', '1'))]
settings = {'quarterLength': 1.0, 'method': 'ffill'}
actual = FilterByOffsetIndexer(in_val, settings).run()
self.assertEqual(actual.index[0], 4.0)
self.assertListEqual(list(actual.index), [4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0])
self.assertEqual(len(actual.columns), 2)
def test_init_4(self):
# when the inputted indices have no parts
in_val = []
setts = {u'quarterLength': 0.5}
self.assertRaises(RuntimeError, FilterByOffsetIndexer, in_val, setts)
try:
FilterByOffsetIndexer(in_val, setts)
except RuntimeError as run_err:
self.assertEqual(FilterByOffsetIndexer._ZERO_PART_ERROR, run_err.args[0])
def test_init_3(self):
# when the specified quarterLength is less than 0.001
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.4, 1.1, 2.1])]
setts = {u'void setting': u'just so void', u'quarterLength': 0.0003}
self.assertRaises(RuntimeError, FilterByOffsetIndexer, in_val, setts)
try:
FilterByOffsetIndexer(in_val, setts)
except RuntimeError as run_err:
self.assertEqual(FilterByOffsetIndexer._QLENGTH_TOO_SMALL_ERROR, run_err.args[0])
def test_init_2(self):
# when there is no quarterLength specified
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.4, 1.1, 2.1])]
setts = {u'void setting': u'just so void'}
self.assertRaises(RuntimeError, FilterByOffsetIndexer, in_val, setts)
try:
FilterByOffsetIndexer(in_val, setts)
except RuntimeError as run_err:
self.assertEqual(FilterByOffsetIndexer._NO_QLENGTH_ERROR, run_err.args[0])
def test_offset_xparts_4(self):
# irregular offset interval to 1.0; 3 parts, same offsets
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.4, 1.1, 2.9], name=('N', '0')),
pandas.Series(['q', 'w', 'e', 'r'], index=[0.0, 0.3, 1.4, 2.6], name=('N', '1')),
pandas.Series(['t', 'a', 'l', 'l'], index=[0.0, 0.2, 1.9, 2.555], name=('N', '2'))]
expected = {'0': pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 1.0, 2.0, 3.0]),
'1': pandas.Series(['q', 'w', 'e', 'r'], index=[0.0, 1.0, 2.0, 3.0]),
'2': pandas.Series(['t', 'a', 'l', 'l'], index=[0.0, 1.0, 2.0, 3.0])}
expected = pandas.DataFrame(expected)
offset_interval = 1.0
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(len(expected.columns), len(actual.columns))
expected = expected.fillna(value='None') # avoid having to use isnan()
actual = actual.fillna(value='None')
for partname in expected.columns:
self.assertSequenceEqual(list(expected[partname].values), list(actual[partname].values))
self.assertSequenceEqual(list(expected[partname].index), list(actual[partname].index))
def test_offset_xparts_0b(self):
# 0 length, many parts, but one part has stuff
in_val = [pandas.Series(name=('N', '0')), pandas.Series(['a', 'b', 'c'], name=('N', '1'),
index=[0.0, 0.5, 1.0]),
pandas.Series(name=('N', '2')), pandas.Series(name=('N', '3'))]
expected = {str(i): pandas.Series() for i in [0, 2, 3]}
#print(str(expected)) # DEBUG
expected['1'] = pandas.Series(['a', 'b', 'c'], index=[0.0, 0.5, 1.0])
#print(str(expected)) # DEBUG
expected = pandas.DataFrame(expected)
offset_interval = 0.5
ind = FilterByOffsetIndexer(in_val, {u'quarterLength': offset_interval})
actual = ind.run()['offset.FilterByOffsetIndexer']
self.assertEqual(len(expected.columns), len(actual.columns))
expected = expected.fillna(value='None') # avoid having to use isnan()
actual = actual.fillna(value='None')
for partname in expected.columns:
self.assertSequenceEqual(list(expected[partname].values), list(actual[partname].values))
self.assertSequenceEqual(list(expected[partname].index), list(actual[partname].index))
def test_init_1(self):
# ensure that __init__() properly sets the "method" setting
in_val = [pandas.Series(['a', 'b', 'c', 'd'], index=[0.0, 0.4, 1.1, 2.1])]
settings = {'quarterLength': 0.5, 'method': 'silly'}
ind = FilterByOffsetIndexer(in_val, settings)
self.assertEqual('silly', ind._settings['method'])