def test_getZScores(self):
"""Data taken from C++ test"""
values = [12,13,9,18,7,9,14,16,10,12,7,13,14,19,10,16,12,16,19,11]
arr = numpy.array(values,dtype=numpy.float64)
zscore = Stats.getZscore(arr)
self.assertAlmostEqual(1.63977, zscore[4], places = 4)
self.assertAlmostEqual(0.32235, zscore[6], places = 4)
modZ = Stats.getModifiedZscore(arr)
self.assertAlmostEqual(1.23658, modZ[4], places = 4)
self.assertAlmostEqual(0.33725, modZ[6], places = 4)
def test_getZScores(self):
"""Data taken from C++ test"""
values = [
12, 13, 9, 18, 7, 9, 14, 16, 10, 12, 7, 13, 14, 19, 10, 16, 12, 16,
19, 11
]
arr = numpy.array(values, dtype=numpy.float64)
zscore = Stats.getZscore(arr)
self.assertAlmostEqual(1.63977, zscore[4], places=4)
self.assertAlmostEqual(0.32235, zscore[6], places=4)
modZ = Stats.getModifiedZscore(arr)
self.assertAlmostEqual(1.23658, modZ[4], places=4)
self.assertAlmostEqual(0.33725, modZ[6], places=4)
def PyExec(self):
in_ws = mtd[self.getPropertyValue('InputWorkspace')]
out_ws_name = self.getPropertyValue('OutputWorkspace')
out_ws = ms.CreateEmptyTableWorkspace(OutputWorkspace=out_ws_name)
out_ws.addColumn('str', 'statistic')
stats = {
'standard_deviation': dict(),
'maximum': dict(),
'minimum': dict(),
'mean': dict(),
'median': dict(),
}
for name in in_ws.getColumnNames():
try:
col_stats = _stats_to_dict(Stats.getStatistics(np.array([float(v) for v in in_ws.column(name)])))
for statname in stats:
stats[statname][name] = col_stats[statname]
out_ws.addColumn('float', name)
except ValueError:
logger.notice('Column \'%s\' is not numerical, skipping' % name)
for name, stat in iteritems(stats):
stat1 = dict(stat)
stat1['statistic'] = name
out_ws.addRow(stat1)
self.setProperty('OutputWorkspace', out_ws)
def PyExec(self):
in_ws = mtd[self.getPropertyValue('InputWorkspace')]
out_ws_name = self.getPropertyValue('OutputWorkspace')
out_ws = ms.CreateEmptyTableWorkspace(OutputWOrkspace=out_ws_name)
out_ws.addColumn('str', 'statistic')
stats = {
'standard_deviation': dict(),
'maximum': dict(),
'minimum': dict(),
'mean': dict(),
'median': dict(),
}
for name in in_ws.getColumnNames():
try:
col_stats = _stats_to_dict(
Stats.getStatistics(
np.array([float(v) for v in in_ws.column(name)])))
for statname in stats.keys():
stats[statname][name] = col_stats[statname]
out_ws.addColumn('float', name)
except ValueError:
logger.notice('Column \'%s\' is not numerical, skipping' %
name)
for name, stat in stats.items():
stat1 = dict(stat)
stat1['statistic'] = name
out_ws.addRow(stat1)
self.setProperty('OutputWorkspace', out_ws_name)
def test_getStatistics_with_floats(self):
data = numpy.array([17.2,18.1,16.5,18.3,12.6])
stats = Stats.getStatistics(data)
self.assertAlmostEqual(16.54, stats.mean, places = 10)
self.assertAlmostEqual(2.0733, stats.standard_deviation, places = 4)
self.assertEquals(12.6, stats.minimum)
self.assertEquals(18.3, stats.maximum)
self.assertEquals(17.2, stats.median)
data = numpy.sort(data)
stats = Stats.getStatistics(data, sorted=True)
self.assertAlmostEqual(16.54, stats.mean, places = 10)
self.assertAlmostEqual(2.0733, stats.standard_deviation, places = 4)
self.assertEquals(12.6, stats.minimum)
self.assertEquals(18.3, stats.maximum)
self.assertEquals(17.2, stats.median)
def test_getStatistics_with_floats(self):
data = numpy.array([17.2, 18.1, 16.5, 18.3, 12.6])
stats = Stats.getStatistics(data)
self.assertAlmostEqual(16.54, stats.mean, places=10)
self.assertAlmostEqual(2.0733, stats.standard_deviation, places=4)
self.assertEquals(12.6, stats.minimum)
self.assertEquals(18.3, stats.maximum)
self.assertEquals(17.2, stats.median)
data = numpy.sort(data)
stats = Stats.getStatistics(data, sorted=True)
self.assertAlmostEqual(16.54, stats.mean, places=10)
self.assertAlmostEqual(2.0733, stats.standard_deviation, places=4)
self.assertEquals(12.6, stats.minimum)
self.assertEquals(18.3, stats.maximum)
self.assertEquals(17.2, stats.median)
def test_getZScores(self):
"""Data taken from C++ test"""
values = [12,13,9,18,7,9,14,16,10,12,7,13,14,19,10,16,12,16,19,11]
arr = numpy.array(values,dtype=numpy.float64)
zscore = Stats.getZscore(arr)
self.assertAlmostEqual(1.63977, zscore[4], places = 4)
self.assertAlmostEqual(0.32235, zscore[6], places = 4)
modZ = Stats.getModifiedZscore(arr)
self.assertAlmostEqual(1.23658, modZ[4], places = 4)
self.assertAlmostEqual(0.33725, modZ[6], places = 4)
# Test the sorted argument still works. Remove this when the function is removed
# sorted=True only ever affected the order
zscore = Stats.getZscore(arr, sorted=True)
self.assertAlmostEqual(1.63977, zscore[4], places = 4)
self.assertAlmostEqual(0.32235, zscore[6], places = 4)
def test_getZScores(self):
"""Data taken from C++ test"""
values = [12,13,9,18,7,9,14,16,10,12,7,13,14,19,10,16,12,16,19,11]
arr = numpy.array(values,dtype=numpy.float64)
zscore = Stats.getZscore(arr)
self.assertAlmostEqual(1.63977, zscore[4], places = 4)
self.assertAlmostEqual(0.32235, zscore[6], places = 4)
modZ = Stats.getModifiedZscore(arr)
self.assertAlmostEqual(1.23658, modZ[4], places = 4)
self.assertAlmostEqual(0.33725, modZ[6], places = 4)
# Test the sorted argument still works. Remove this when the function is removed
# sorted=True only ever affected the order
zscore = Stats.getZscore(arr, sorted=True)
self.assertAlmostEqual(1.63977, zscore[4], places = 4)
self.assertAlmostEqual(0.32235, zscore[6], places = 4)
def test_getMoments(self):
mean = 5.
sigma = 4.
deltaX = .2
numX = 200
# calculate to have same number of points left and right of function
offsetX = mean - (.5 * deltaX * float(numX))
# variance about origin
expVar = mean * mean + sigma * sigma
# skew about origin
expSkew = mean * mean * mean + 3. * mean * sigma * sigma
# x-values to try out
indep = numpy.arange(numX, dtype=numpy.float64)
indep = indep * deltaX + offsetX
# y-values
# test different type
depend = numpy.arange(numX, dtype=numpy.int32)
self.assertRaises(ValueError, Stats.getMomentsAboutOrigin, indep,
depend)
# now correct y values
weightedDiff = (indep - mean) / sigma
depend = numpy.exp(-0.5 * weightedDiff *
weightedDiff) / sigma / math.sqrt(2. * math.pi)
aboutOrigin = Stats.getMomentsAboutOrigin(indep, depend)
self.assertTrue(isinstance(aboutOrigin, numpy.ndarray))
self.assertEquals(4, aboutOrigin.shape[0])
self.assertAlmostEqual(1., aboutOrigin[0], places=4)
self.assertAlmostEqual(mean, aboutOrigin[1], places=4)
self.assertTrue(math.fabs(expVar - aboutOrigin[2]) < 0.001 * expVar)
self.assertTrue(math.fabs(expSkew - aboutOrigin[3]) < 0.001 * expSkew)
aboutMean = Stats.getMomentsAboutMean(indep, depend)
self.assertTrue(isinstance(aboutOrigin, numpy.ndarray))
self.assertEquals(4, aboutOrigin.shape[0])
self.assertAlmostEqual(1., aboutMean[0], places=4)
self.assertAlmostEqual(0., aboutMean[1], places=4)
self.assertTrue(
math.fabs(sigma * sigma - aboutMean[2]) < 0.001 * expVar)
self.assertTrue(math.fabs(0. - aboutMean[3]) < 0.0001 * expSkew)
def test_getMoments(self):
mean = 5.
sigma = 4.
deltaX = .2
numX = 200
# calculate to have same number of points left and right of function
offsetX = mean - (.5 * deltaX * float(numX))
# variance about origin
expVar = mean*mean+sigma*sigma;
# skew about origin
expSkew = mean*mean*mean+3.*mean*sigma*sigma;
# x-values to try out
indep = numpy.arange(numX, dtype=numpy.float64)
indep = indep*deltaX + offsetX
# y-values
# test different type
depend = numpy.arange(numX, dtype=numpy.int32)
self.assertRaises(ValueError, Stats.getMomentsAboutOrigin, indep, depend)
# now correct y values
weightedDiff = (indep-mean)/sigma
depend = numpy.exp(-0.5*weightedDiff*weightedDiff)/sigma/math.sqrt(2.*math.pi)
aboutOrigin = Stats.getMomentsAboutOrigin(indep, depend)
self.assertTrue(isinstance(aboutOrigin, numpy.ndarray))
self.assertEquals(4, aboutOrigin.shape[0])
self.assertAlmostEqual(1., aboutOrigin[0], places=4)
self.assertAlmostEqual(mean, aboutOrigin[1], places=4)
self.assertTrue(math.fabs(expVar - aboutOrigin[2]) < 0.001*expVar)
self.assertTrue(math.fabs(expSkew - aboutOrigin[3]) < 0.001*expSkew)
aboutMean = Stats.getMomentsAboutMean(indep, depend)
self.assertTrue(isinstance(aboutOrigin, numpy.ndarray))
self.assertEquals(4, aboutOrigin.shape[0])
self.assertAlmostEqual(1., aboutMean[0], places=4)
self.assertAlmostEqual(0., aboutMean[1], places=4)
self.assertTrue(math.fabs(sigma*sigma - aboutMean[2]) < 0.001*expVar)
self.assertTrue(math.fabs(0. - aboutMean[3]) < 0.0001*expSkew)
def PyExec(self):
in_ws = mtd[self.getPropertyValue('InputWorkspace')]
indices_list = self.getPropertyValue('ColumnIndices')
out_ws_name = self.getPropertyValue('OutputWorkspace')
column_names = in_ws.getColumnNames()
# If column indices are not provided, then default to _ALL_ columns
if len(indices_list) > 0:
indices_list = [int(x) for x in indices_list.split(',')]
else:
indices_list = range(len(column_names))
out_ws = ms.CreateEmptyTableWorkspace(OutputWorkspace=out_ws_name)
out_ws.addColumn('str', 'Statistic')
stats = collections.OrderedDict([
('StandardDev', collections.OrderedDict()),
('Minimum', collections.OrderedDict()),
('Median', collections.OrderedDict()),
('Maximum', collections.OrderedDict()),
('Mean', collections.OrderedDict()),
])
for index in indices_list:
column_name = column_names[index]
try:
column_data = np.array([float(v) for v in in_ws.column(index)])
col_stats = _stats_to_dict(Stats.getStatistics(column_data))
for stat_name in stats:
stats[stat_name][column_name] = col_stats[stat_name]
out_ws.addColumn('float', column_name)
except RuntimeError:
logger.notice('Column \'%s\' is not numerical, skipping' %
column_name)
except:
logger.notice('Column \'%s\' is not numerical, skipping' %
column_name)
for index, stat_name in iteritems(stats):
stat = collections.OrderedDict(stat_name)
stat['Statistic'] = index
out_ws.addRow(stat)
self.setProperty('OutputWorkspace', out_ws)
def PyExec(self):
in_ws = mtd[self.getPropertyValue('InputWorkspace')]
indices_list = self.getPropertyValue('ColumnIndices')
out_ws_name = self.getPropertyValue('OutputWorkspace')
column_names = in_ws.getColumnNames()
# If column indices are not provided, then default to _ALL_ columns
if len(indices_list) > 0:
indices_list = [int(x) for x in indices_list.split(',')]
else:
indices_list = range(len(column_names))
out_ws = ms.CreateEmptyTableWorkspace(OutputWorkspace=out_ws_name)
out_ws.addColumn('str', 'Statistic')
stats = collections.OrderedDict([
('StandardDev', collections.OrderedDict()),
('Minimum', collections.OrderedDict()),
('Median', collections.OrderedDict()),
('Maximum', collections.OrderedDict()),
('Mean', collections.OrderedDict()),
])
for index in indices_list:
column_name = column_names[index]
try:
column_data = np.array([float(v) for v in in_ws.column(index)])
col_stats = _stats_to_dict(Stats.getStatistics(column_data))
for stat_name in stats:
stats[stat_name][column_name] = col_stats[stat_name]
out_ws.addColumn('float', column_name)
except RuntimeError:
logger.notice('Column \'%s\' is not numerical, skipping' % column_name)
except:
logger.notice('Column \'%s\' is not numerical, skipping' % column_name)
for index, stat_name in iteritems(stats):
stat = collections.OrderedDict(stat_name)
stat['Statistic'] = index
out_ws.addRow(stat)
self.setProperty('OutputWorkspace', out_ws)