def test_not(self):
data = numpy.zeros(100, dtype=[('mass', int), ('velocity', int)])
data['mass'] = numpy.random.choice([0, 1, 2], size=100)
ans = data['mass'] >= 1
q = Query("not mass < 1")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
q = Query("not (mass < 1)")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_compound_or(self):
data = numpy.zeros(100, dtype=[('mass', int), ('velocity', int)])
data['mass'] = numpy.random.choice([0, 1, 2], size=100)
data['velocity'] = numpy.random.choice([0, 1, 2], size=100)
ans = (data['mass'] == 0) | (data['velocity'] != 1)
q = Query("mass == 0 or velocity != 1")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
q = Query("(mass == 0) or (velocity != 1)")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_compound_and(self):
data = numpy.zeros(100, dtype=[('mass', float), ('velocity', float)])
data['mass'] = numpy.random.random(size=100)
data['velocity'] = numpy.random.random(size=100)
q = Query("mass < 0.5 and velocity > 0.5")
sliced = data[q.get_mask(data)]
self.assertTrue(numpy.alltrue(sliced['mass'] < 0.5))
self.assertTrue(numpy.alltrue(sliced['velocity'] > 0.5))
q = Query("(mass < 0.5) and (velocity > 0.5)")
sliced = data[q.get_mask(data)]
self.assertTrue(numpy.alltrue(sliced['mass'] < 0.5))
self.assertTrue(numpy.alltrue(sliced['velocity'] > 0.5))
def test_missing_op(self):
data = numpy.zeros(5, dtype=[('mass', 'f8')])
data['mass'][0] = numpy.nan
with pytest.raises(SelectionError):
q = Query("mass isnot 4)")
m = q.get_mask(data)
def test_eval_fail(self):
data = numpy.zeros(5, dtype=[('mass', 'f8')])
data['mass'][0] = numpy.nan
with pytest.raises(SelectionError):
q = Query("mass < missing_func(0.3)")
m = q.get_mask(data)
def test_indexing_out_of_range(self):
data = numpy.zeros(10, dtype=[('Position', (float, 3))])
data['Position'] = numpy.random.random(size=(10, 3))
with pytest.raises(SelectionError):
q = Query("Position[:,4] is nan")
m = q.get_mask(data)
def test_indexing_single_dimension(self):
data = numpy.zeros(5, dtype=[('mass', 'f8')])
data['mass'][0] = numpy.nan
with pytest.raises(SelectionError):
q = Query("mass[:,0] is nan")
m = q.get_mask(data)
def test_missing_column(self):
data = numpy.zeros(5, dtype=[('mass', 'f8')])
data['mass'][0] = numpy.nan
with pytest.raises(SelectionError):
q = Query("velocity is nan")
m = q.get_mask(data)
def test_slice_index_fail(self):
data = numpy.zeros(100, dtype=[('Position', (float, 3))])
data['Position'] = numpy.random.random(size=(100, 3))
# output index must be 1D or same dimension as input
with pytest.raises(SelectionError):
q = Query("Position[:,:2] < 0.5")
mask = q.get_mask(data)
def test_is_not(self):
data = numpy.zeros(5, dtype=[('mass', 'int')])
data['mass'][0] = 1
q = Query("mass is not 0")
m = q.get_mask(data)
self.assertTrue(m.sum() == 1)
self.assertTrue(m[0] == True)
def test_compound_log(self):
data = numpy.zeros(100, dtype=[('mass', float)])
data['mass'] = numpy.random.random(size=100)
ans = (data['mass'] < 0.5 * numpy.log(2.))
q = Query("mass < 0.5*log(2)")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_minus(self):
data = numpy.zeros(100, dtype=[('mass', float)])
data['mass'] = numpy.random.random(size=100)
ans = (data['mass'] < 0.25)
q = Query("mass < 0.5-0.125-0.125")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_nan(self):
data = numpy.zeros(5, dtype=[('mass', 'f8')])
data['mass'][0] = numpy.nan
q = Query("mass is nan")
m = q.get_mask(data)
self.assertTrue(m.sum() == 1)
self.assertTrue(m[0] == True)
def test_index(self):
data = numpy.zeros(100, dtype=[('Position', (float, 3))])
data['Position'] = numpy.random.random(size=(100, 3))
ans = (data['Position'][:, 0] < 0.5) & (data['Position'][:, -1] >= 0.7)
q = Query("Position[:,0] < 0.5 and Position[:,-1] >= 0.7")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_column_log10(self):
data = numpy.zeros(100, dtype=[('mass', float)])
data['mass'] = numpy.random.random(size=100)
ans = numpy.log10(data['mass']) > -0.5
q = Query("log10(mass) > -0.5")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_not_inf(self):
data = numpy.zeros(5, dtype=[('mass', 'f8')])
data['mass'][0] = numpy.inf
q = Query("mass is not inf")
m = q.get_mask(data)
self.assertTrue(m.sum() == len(data) - 1)
self.assertTrue(m[0] == False)
def test_exp(self):
data = numpy.zeros(100, dtype=[('mass', float)])
data['mass'] = numpy.random.random(size=100)
ans = (data['mass'] < numpy.exp(-0.9))
q = Query("mass < exp(-0.9))")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))
def test_gt(self):
data = numpy.zeros(5, dtype=[('mass', 'int')])
data['mass'][0] = 1
data['mass'][1] = 2
q = Query("mass > 1")
m = q.get_mask(data)
self.assertTrue(m.sum() == 1)
self.assertTrue(m[1] == True)
def test_lt(self):
data = numpy.zeros(5, dtype=[('mass', 'int')])
data['mass'][0] = 1
data['mass'][1] = 2
q = Query("mass < 1")
m = q.get_mask(data)
self.assertTrue(m.sum() == len(data) - 2)
self.assertTrue((m[0] == False) and (m[1] == False))
def test_compound_index(self):
data = numpy.zeros(100,
dtype=[('Position', (float, 3)), ('velocity', int)])
data['Position'] = numpy.random.random(size=(100, 3))
data['velocity'] = numpy.random.choice([0, 1, 2], size=100)
ans = (data['Position'][:, 1] < 0.5) | (data['velocity'] == 2)
q = Query("Position[:,1] < 0.5 or velocity == 2")
mask = q.get_mask(data)
self.assertTrue(numpy.all(ans == mask))