def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
# Create the output dir if it doesn't already exist.
output_dir = opts.output_dir
try:
create_dir(output_dir)
except:
option_parser.error("Could not create or access output directory "
"specified with the -o/--output_dir option.")
otu_table_fp = opts.otu_table_fp
table = load_table(otu_table_fp)
estimator = ObservationRichnessEstimator(table,
Chao1MultinomialPointEstimator)
results = estimator(opts.min, opts.max, opts.num_steps,
opts.confidence_level)
out_fp = join(output_dir, 'estimates_table.txt')
with open(out_fp, 'w') as out_f:
results.toTable(out_f)
def setUp(self):
"""Define some sample data that will be used by the tests."""
# Single sample, 6 observations, one of which isn't observed in sample.
self.biom_table1 = parse_biom_table(biom_table_str1)
self.estimator1 = ObservationRichnessEstimator(
self.biom_table1, Chao1MultinomialPointEstimator)
def setUp(self):
"""Define some sample data that will be used by the tests."""
# Single sample, 6 observations, one of which isn't observed in sample.
self.biom_table1 = parse_biom_table(biom_table_str1)
self.estimator1 = ObservationRichnessEstimator(self.biom_table1,
Chao1MultinomialPointEstimator)
class ObservationRichnessEstimatorTests(TestCase):
"""Tests for the ObservationRichnessEstimator class."""
def setUp(self):
"""Define some sample data that will be used by the tests."""
# Single sample, 6 observations, one of which isn't observed in sample.
self.biom_table1 = parse_biom_table(biom_table_str1)
self.estimator1 = ObservationRichnessEstimator(
self.biom_table1, Chao1MultinomialPointEstimator)
def test_constructor(self):
"""Test instantiating an ObservationRichnessEstimator."""
self.assertTrue(
isinstance(self.estimator1, ObservationRichnessEstimator))
def test_constructor_empty_table(self):
"""Test instantiating an estimator with an empty table."""
empty_table = Table(array([]), [], [])
self.assertRaises(EmptyTableError, ObservationRichnessEstimator,
empty_table, Chao1MultinomialPointEstimator)
def test_getSampleCount(self):
"""Test estimator returns correct number of samples."""
self.assertEqual(self.estimator1.getSampleCount(), 1)
def test_call_interpolate(self):
"""Test __call__ computes correct estimates (interpolation)."""
# Verified with iNEXT (http://glimmer.rstudio.com/tchsieh/inext/).
# SE estimates differ because they use a different technique. SE
# estimates have been verified against values in Colwell 2012 instead
# (in separate unit tests).
# Just reference.
obs = self.estimator1(start=15, stop=15, num_steps=1)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(
obs.getEstimates('S1'),
[(15, 5, 0.674199862463, 3.67859255119, 6.32140744881)])
# start=1 and reference.
obs = self.estimator1(start=1, stop=1, num_steps=1)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(
obs.getEstimates('S1'),
[(1, 1.0, 0.250252397843, 0.509514313183, 1.49048568682),
(15, 5, 0.674199862463, 3.67859255119, 6.32140744881)])
# Points in between start=1 and reference.
obs = self.estimator1(start=1, stop=15, num_steps=3)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(
obs.getEstimates('S1'),
[(1, 1.0, 0.250252397843, 0.509514313183, 1.49048568682),
(5, 3.40326340326, 0.655024590447, 2.119438797, 4.68708800953),
(9, 4.4001998002, 0.680106580075, 3.0672153976, 5.7331842028),
(13, 4.85714285714, 0.665379090563, 3.55302380357, 6.16126191071),
(15, 5, 0.674199862463, 3.67859255119, 6.32140744881)])
def test_call_extrapolate(self):
"""Test __call__ computes correct estimates (extrapolation)."""
# Verified with iNEXT. Differs slightly from their output because
# they've slightly modified Colwell 2012 equation 9, and we're using
# the original one. SE estimates differ because they use a different
# technique. SE estimates have been verified against values in Colwell
# 2012 instead (in separate unit tests).
obs = self.estimator1(start=15, stop=30, num_steps=1)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(
obs.getEstimates('S1'),
[(15, 5, 0.674199862463, 3.67859255119, 6.32140744881),
(30, 5.4415544562981095, 1.073911829557642, 3.33672594779,
7.5463829648)])
obs = self.estimator1(start=20, stop=30, num_steps=2)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(
obs.getEstimates('S1'),
[(15, 5, 0.674199862463, 3.67859255119, 6.32140744881),
(20, 5.2555272427983537, 0.77331345626875192, 3.73986071975,
6.77119376585),
(25, 5.38046614197245, 0.93220670591157662, 3.55337457224,
7.20755771171),
(30, 5.4415544562981095, 1.073911829557642, 3.33672594779,
7.5463829648)])
def test_get_points_to_estimate_invalid_input(self):
"""Raises an error on invalid input."""
# Invalid min.
self.assertRaises(ValueError, self.estimator1._get_points_to_estimate,
5, 0, 10, 1)
# Invalid num_steps.
self.assertRaises(ValueError, self.estimator1._get_points_to_estimate,
5, 1, 10, 0)
# max < min.
self.assertRaises(ValueError, self.estimator1._get_points_to_estimate,
5, 1, -1, 1)
def test_get_points_to_estimate(self):
"""Correctly calculates estimation points given range parameters."""
# Ref in range.
obs = self.estimator1._get_points_to_estimate(4, 1, 5, 4)
self.assertEqual(obs, [1, 2, 3, 4, 5])
# Ref not in range.
obs = self.estimator1._get_points_to_estimate(4, 5, 10, 2)
self.assertEqual(obs, [4, 5, 7, 9])
# stop not supplied.
obs = self.estimator1._get_points_to_estimate(5, 5, num_steps=2)
self.assertEqual(obs, [5, 17, 29])
class ObservationRichnessEstimatorTests(TestCase):
"""Tests for the ObservationRichnessEstimator class."""
def setUp(self):
"""Define some sample data that will be used by the tests."""
# Single sample, 6 observations, one of which isn't observed in sample.
self.biom_table1 = parse_biom_table(biom_table_str1)
self.estimator1 = ObservationRichnessEstimator(self.biom_table1,
Chao1MultinomialPointEstimator)
def test_constructor(self):
"""Test instantiating an ObservationRichnessEstimator."""
self.assertTrue(isinstance(self.estimator1,
ObservationRichnessEstimator))
def test_constructor_empty_table(self):
"""Test instantiating an estimator with an empty table."""
empty_table = Table(array([]), [], [])
self.assertRaises(EmptyTableError, ObservationRichnessEstimator,
empty_table, Chao1MultinomialPointEstimator)
def test_getSampleCount(self):
"""Test estimator returns correct number of samples."""
self.assertEqual(self.estimator1.getSampleCount(), 1)
def test_call_interpolate(self):
"""Test __call__ computes correct estimates (interpolation)."""
# Verified with iNEXT (http://glimmer.rstudio.com/tchsieh/inext/).
# SE estimates differ because they use a different technique. SE
# estimates have been verified against values in Colwell 2012 instead
# (in separate unit tests).
# Just reference.
obs = self.estimator1(start=15, stop=15, num_steps=1)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(obs.getEstimates('S1'),
[(15, 5, 0.674199862463, 3.67859255119, 6.32140744881)])
# start=1 and reference.
obs = self.estimator1(start=1, stop=1, num_steps=1)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(obs.getEstimates('S1'),
[(1, 1.0, 0.250252397843, 0.509514313183, 1.49048568682),
(15, 5, 0.674199862463, 3.67859255119, 6.32140744881)])
# Points in between start=1 and reference.
obs = self.estimator1(start=1, stop=15, num_steps=3)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(obs.getEstimates('S1'),
[(1, 1.0, 0.250252397843, 0.509514313183, 1.49048568682),
(5, 3.40326340326, 0.655024590447, 2.119438797,
4.68708800953),
(9, 4.4001998002, 0.680106580075, 3.0672153976, 5.7331842028),
(13, 4.85714285714, 0.665379090563, 3.55302380357,
6.16126191071),
(15, 5, 0.674199862463, 3.67859255119, 6.32140744881)])
def test_call_extrapolate(self):
"""Test __call__ computes correct estimates (extrapolation)."""
# Verified with iNEXT. Differs slightly from their output because
# they've slightly modified Colwell 2012 equation 9, and we're using
# the original one. SE estimates differ because they use a different
# technique. SE estimates have been verified against values in Colwell
# 2012 instead (in separate unit tests).
obs = self.estimator1(start=15, stop=30, num_steps=1)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(obs.getEstimates('S1'),
[(15, 5, 0.674199862463, 3.67859255119, 6.32140744881),
(30, 5.4415544562981095, 1.073911829557642, 3.33672594779,
7.5463829648)])
obs = self.estimator1(start=20, stop=30, num_steps=2)
self.assertEqual(obs.getSampleCount(), 1)
self.assertFloatEqual(obs.getEstimates('S1'),
[(15, 5, 0.674199862463, 3.67859255119, 6.32140744881),
(20, 5.2555272427983537, 0.77331345626875192, 3.73986071975,
6.77119376585),
(25, 5.38046614197245, 0.93220670591157662, 3.55337457224,
7.20755771171),
(30, 5.4415544562981095, 1.073911829557642, 3.33672594779,
7.5463829648)])
def test_get_points_to_estimate_invalid_input(self):
"""Raises an error on invalid input."""
# Invalid min.
self.assertRaises(ValueError, self.estimator1._get_points_to_estimate,
5, 0, 10, 1)
# Invalid num_steps.
self.assertRaises(ValueError, self.estimator1._get_points_to_estimate,
5, 1, 10, 0)
# max < min.
self.assertRaises(ValueError, self.estimator1._get_points_to_estimate,
5, 1, -1, 1)
def test_get_points_to_estimate(self):
"""Correctly calculates estimation points given range parameters."""
# Ref in range.
obs = self.estimator1._get_points_to_estimate(4, 1, 5, 4)
self.assertEqual(obs, [1, 2, 3, 4, 5])
# Ref not in range.
obs = self.estimator1._get_points_to_estimate(4, 5, 10, 2)
self.assertEqual(obs, [4, 5, 7, 9])
# stop not supplied.
obs = self.estimator1._get_points_to_estimate(5, 5, num_steps=2)
self.assertEqual(obs, [5, 17, 29])
