def test_phases(self):
"Test indices"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator = ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
# Test cold/neutral
cold = series.cold
assert_equal(cold._mask,
[ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0,
0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,])
neutral = series.neutral
assert_equal(neutral._mask,
[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1,
1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1,
1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,])
warm = series.warm
assert_equal(warm._mask,
[ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0,
0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1,])
def test_pickling(self):
"Tests pickling"
import cPickle
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
pickled = cPickle.loads(cPickle.dumps(series))
assert_equal(series, pickled)
assert_equal(pickled.ensoindicator, pickled.ensoindicator)
assert_equal(series.ensoindices, pickled.ensoindices)
def test_pickling(self):
"Tests pickling"
import cPickle
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator = ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
pickled = cPickle.loads(cPickle.dumps(series))
assert_equal(series, pickled)
assert_equal(pickled.ensoindicator, pickled.ensoindicator)
assert_equal(series.ensoindices, pickled.ensoindices)
def test_creation(self):
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
local_ensoi = series.ensoindicator
assert_equal(local_ensoi, ensoi)
assert_equal(local_ensoi.optinfo, ensoi.optinfo)
def test_apply_on_phase(self):
"Test apply on phase"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
func = ma.mean
_cold = func(series[(series.ensoindices == -1).filled()])
_neut = func(series[(series.ensoindices == 0).filled()])
_warm = func(series[(series.ensoindices == +1).filled()])
_glob = func(series)
control = ma.array([(_cold, _neut, _warm, _glob)],
dtype=[('cold', float), ('neutral', float),
('warm', float), ('global', float)])
control['warm'] = masked
result = apply_on_phase(series, func)
def test_reshape(self):
"Test reshape"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator = ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
#
series = series.reshape(1, -1)
assert_equal(series.ensoindicator.shape, series.shape)
self.failUnless(series.ensoindicator.shape != ensoi.shape)
assert_equal(series.ensoindicator, ensoi.reshape(1, -1))
#
series.shape = (-1, 1)
assert_equal(series.ensoindicator.shape, series.shape)
self.failUnless(series.ensoindicator.shape != ensoi.shape)
assert_equal(series.ensoindicator, ensoi.reshape(-1, 1))
def test_apply_on_phase(self):
"Test apply on phase"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator = ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
func = ma.mean
_cold = func(series[(series.ensoindices==-1).filled()])
_neut = func(series[(series.ensoindices==0).filled()])
_warm = func(series[(series.ensoindices==+1).filled()])
_glob = func(series)
control = ma.array([(_cold, _neut, _warm, _glob)],
dtype=[('cold',float),('neutral',float),
('warm',float),('global',float)])
control['warm'] = masked
result = apply_on_phase(series, func)
def test_creation_shifted(self):
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates - 24,
ensoindicator=ensoi)
local_enso = series.ensoindicator
assert_equal(local_enso._mask, [1] * 24 + [0] * (len(ensoi) - 24))
assert_equal(local_enso.dates, series.dates)
assert_equal(local_enso.compressed(), ensoi[:(len(ensoi) - 24)])
assert_equal(local_enso.optinfo, ensoi.optinfo)
def test_indices(self):
"Test indices"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator = ensoi)
# Try without initializiation: it should fail
try:
indc = series.ensoindices
except ValueError:
pass
# So, initialize it:
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
indc = series.ensoindices
assert_equal(indc,
[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-1,
-1,-1,-1,-1,-1, 0, 0,-1,-1,-1,-1,-1,
-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0,+1,+1,
+1,+1,+1,+1,+1,+1,+1, 0, 0, 0, 0, 0,])
def test_indices_convert(self):
"Test the conversion of ensoindices from one frequency to another."
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator = ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
control = ts.time_series([ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-1,
-1,-1,-1,-1,-1, 0, 0,-1,-1,-1,-1,-1,
-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0,+1,+1,
+1,+1,+1,+1,+1,+1,+1, 0, 0, 0, 0, 0,],
dates=ensoi.dates)
assert_equal(series.ensoindices, control)
# Conversion 'M' to 'D'
dseries = series.convert('D')
assert_equal(dseries.ensoindices,
ts.lib.backward_fill(control.convert('D')))
# Conversion 'M' to 'A'
aseries = series.convert('A', func=ma.mean)
assert_equal(aseries.ensoindices,
mode(control.convert('A'), axis=1)[0].squeeze())
def test_getitem_with_string(self):
"Test getitem w/ a string"
ensoi = self.ensoi
series = ClimateSeries(zip(
np.random.rand(len(ensoi)),
np.random.rand(len(ensoi)),
),
dtype=[('A', float), ('B', float)],
dates=ensoi._dates,
ensoindicator=ensoi)
# With field
series_A = series['A']
assert_equal(series_A.ensoindicator, ensoi)
assert_equal(series['B'].ensoindicator, ensoi)
# With date
series_first = series['1980-01':'1981-01']
assert_equal(series_first, series[:12])
def test_reshape(self):
"Test reshape"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
#
series = series.reshape(1, -1)
assert_equal(series.ensoindicator.shape, series.shape)
self.failUnless(series.ensoindicator.shape != ensoi.shape)
assert_equal(series.ensoindicator, ensoi.reshape(1, -1))
#
series.shape = (-1, 1)
assert_equal(series.ensoindicator.shape, series.shape)
self.failUnless(series.ensoindicator.shape != ensoi.shape)
assert_equal(series.ensoindicator, ensoi.reshape(-1, 1))
def test_indices(self):
"Test indices"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
# Try without initializiation: it should fail
try:
indc = series.ensoindices
except ValueError:
pass
# So, initialize it:
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
indc = series.ensoindices
assert_equal(indc, [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
0,
0,
-1,
-1,
-1,
-1,
-1,
-1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
+1,
+1,
+1,
+1,
+1,
+1,
+1,
+1,
+1,
0,
0,
0,
0,
0,
])
def test_phases(self):
"Test indices"
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
# Test cold/neutral
cold = series.cold
assert_equal(cold._mask, [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
1,
1,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
])
neutral = series.neutral
assert_equal(neutral._mask, [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
0,
0,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
])
warm = series.warm
assert_equal(warm._mask, [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
])
def test_indices_convert(self):
"Test the conversion of ensoindices from one frequency to another."
ensoi = self.ensoi
series = ClimateSeries(np.random.rand(len(ensoi)),
dates=ensoi._dates,
ensoindicator=ensoi)
series.set_ensoindices(minimum_size=5, reference_season='NDJ')
control = ts.time_series([
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
0,
0,
-1,
-1,
-1,
-1,
-1,
-1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
+1,
+1,
+1,
+1,
+1,
+1,
+1,
+1,
+1,
0,
0,
0,
0,
0,
],
dates=ensoi.dates)
assert_equal(series.ensoindices, control)
# Conversion 'M' to 'D'
dseries = series.convert('D')
assert_equal(dseries.ensoindices,
ts.lib.backward_fill(control.convert('D')))
# Conversion 'M' to 'A'
aseries = series.convert('A', func=ma.mean)
assert_equal(aseries.ensoindices,
mode(control.convert('A'), axis=1)[0].squeeze())
