def test_2d(self):
"Tests median w/ 2D"
(n,p) = (101,30)
x = masked_array(numpy.linspace(-1.,1.,n),)
x[:10] = x[-10:] = masked
z = masked_array(numpy.empty((n,p), dtype=numpy.float_))
z[:,0] = x[:]
idx = numpy.arange(len(x))
for i in range(1,p):
numpy.random.shuffle(idx)
z[:,i] = x[idx]
assert_equal(mmedian(z[:,0]), 0)
assert_equal(mmedian(z), numpy.zeros((p,)))
def test_ranking(self):
x = masked_array([0,1,1,1,2,3,4,5,5,6,])
assert_almost_equal(rank_data(x),[1,3,3,3,5,6,7,8.5,8.5,10])
x[[3,4]] = masked
assert_almost_equal(rank_data(x),[1,2.5,2.5,0,0,4,5,6.5,6.5,8])
assert_almost_equal(rank_data(x,use_missing=True),
[1,2.5,2.5,4.5,4.5,4,5,6.5,6.5,8])
x = masked_array([0,1,5,1,2,4,3,5,1,6,])
assert_almost_equal(rank_data(x),[1,3,8.5,3,5,7,6,8.5,3,10])
x = masked_array([[0,1,1,1,2], [3,4,5,5,6,]])
assert_almost_equal(rank_data(x),[[1,3,3,3,5],[6,7,8.5,8.5,10]])
assert_almost_equal(rank_data(x,axis=1),[[1,3,3,3,5],[1,2,3.5,3.5,5]])
assert_almost_equal(rank_data(x,axis=0),[[1,1,1,1,1],[2,2,2,2,2,]])
def test_winsorization(self):
"Tests the Winsorization of the data."
data = masked_array([ 77, 87, 88,114,151,210,219,246,253,262,
296,299,306,376,428,515,666,1310,2611])
assert_almost_equal(winsorize(data).varu(), 21551.4, 1)
data[5] = masked
winsorized = winsorize(data)
assert_equal(winsorized.mask, data.mask)
def check_cov(self):
"Tests the cov function."
x = masked_array([[1,2,3],[4,5,6]], mask=[[1,0,0],[0,0,0]])
c = cov(x[0])
assert_equal(c, (x[0].anom()**2).sum())
c = cov(x[1])
assert_equal(c, (x[1].anom()**2).sum()/2.)
c = cov(x)
assert_equal(c[1,0], (x[0].anom()*x[1].anom()).sum())
def test_unsorted(self):
"Tests that the data are properly sorted along the dates."
dlist = ['2007-01-%02i' % i for i in (3,2,1)]
data = [10,20,30]
series = time_series(data,dlist)
assert_equal(series._data,[30,20,10])
#
series = TimeSeries(data, dlist)
assert_equal(series._data,[30,20,10])
#
series = TimeSeries(data, dlist, mask=[1,0,0])
assert_equal(series._mask,[0,0,1])
#
data = masked_array([10,20,30],mask=[1,0,0])
series = TimeSeries(data, dlist)
assert_equal(series._mask,[0,0,1])
def test_pickling(self):
"Tests pickling/unpickling"
(series, data, dates) = self.d
import cPickle
series_pickled = cPickle.loads(series.dumps())
assert_equal(series_pickled._dates, series._dates)
assert_equal(series_pickled._data, series._data)
assert_equal(series_pickled._mask, series._mask)
#
data = masked_array(N.matrix(range(10)).T, mask=[1,0,0,0,0]*2)
dates = date_array(start_date=thisday('D'), length=10)
series = time_series(data,dates=dates)
series_pickled = cPickle.loads(series.dumps())
assert_equal(series_pickled._dates, series._dates)
assert_equal(series_pickled._data, series._data)
assert_equal(series_pickled._mask, series._mask)
assert(isinstance(series_pickled._data, N.matrix))
def test_compressed(self):
"Tests compress"
dlist = ['2007-01-%02i' % i for i in range(1,16)]
dates = date_array_fromlist(dlist)
data = masked_array(numeric.arange(15), mask=[1,0,0,0,0]*3, dtype=float_)
series = time_series(data, dlist)
#
keeper = N.array([0,1,1,1,1]*3, dtype=bool_)
c_series = series.compressed()
assert_equal(c_series._data, [1,2,3,4,6,7,8,9,11,12,13,14])
assert_equal(c_series._mask, nomask)
assert_equal(c_series._dates, dates[keeper])
#
series_st = time_series(MA.column_stack((data,data[::-1])),
dates=dates)
c_series = series_st.compressed()
d = [1,2,3,6,7,8,11,12,13]
assert_equal(c_series._data, N.c_[(d,list(reversed(d)))])
assert_equal(c_series._mask, nomask)
assert_equal(c_series._dates, dates[d])
def test_concatenate(self):
"Tests concatenate"
dlist = ['2007-%02i' % i for i in range(1,6)]
dates = date_array_fromlist(dlist)
data = masked_array(numeric.arange(5), mask=[1,0,0,0,0], dtype=float_)
#
ser_1 = time_series(data, dates)
ser_2 = time_series(data, dates=dates+10)
newseries = concatenate_series(ser_1, ser_2)
assert_equal(newseries._data,[0,1,2,3,4,0,0,0,0,0,0,1,2,3,4])
assert_equal(newseries._mask,[1,0,0,0,0]+[1]*5+[1,0,0,0,0])
#
ser_1 = time_series(data, dates)
ser_2 = time_series(data, dates=dates+10)
newseries = concatenate_series(ser_1, ser_2, keep_gap=False)
assert_equal(newseries._data,[0,1,2,3,4,0,1,2,3,4])
assert_equal(newseries._mask,[1,0,0,0,0]+[1,0,0,0,0])
assert newseries.has_missing_dates()
#
ser_2 = time_series(data, dates=dates+3)
newseries = concatenate_series(ser_1, ser_2)
assert_equal(newseries._data,[0,1,2,0,1,2,3,4])
assert_equal(newseries._mask,[1,0,0,1,0,0,0,0])
def test_maskperiod(self):
"Test mask_period"
(series, data, dates) = self.d
series.mask = nomask
(start, end) = ('2007-01-06', '2007-01-12')
mask = mask_period(series, start, end, inside=True, include_edges=True,
inplace=False)
assert_equal(mask._mask, N.array([0,0,0,0,0,1,1,1,1,1,1,1,0,0,0]))
mask = mask_period(series, start, end, inside=True, include_edges=False,
inplace=False)
assert_equal(mask._mask, [0,0,0,0,0,0,1,1,1,1,1,0,0,0,0])
mask = mask_period(series, start, end, inside=False, include_edges=True,
inplace=False)
assert_equal(mask._mask, [1,1,1,1,1,1,0,0,0,0,0,1,1,1,1])
mask = mask_period(series, start, end, inside=False, include_edges=False,
inplace=False)
assert_equal(mask._mask, [1,1,1,1,1,0,0,0,0,0,0,0,1,1,1])
# Now w/ multivariables
data = masked_array(numeric.arange(30).reshape(-1,2), dtype=float_)
series = time_series(data, dates=dates)
mask = mask_period(series, start, end, inside=True, include_edges=True,
inplace=False)
result = N.array([0,0,0,0,0,1,1,1,1,1,1,1,0,0,0])
assert_equal(mask._mask, result.repeat(2).reshape(-1,2))
def test_trimmed_stde(self):
"Tests the trimmed mean standard error."
data = masked_array([ 77, 87, 88,114,151,210,219,246,253,262,
296,299,306,376,428,515,666,1310,2611])
assert_almost_equal(trimmed_stde(data,0.2), 56.1, 1)
def test_trimmedmean(self):
"Tests the trimmed mean."
data = masked_array([ 77, 87, 88,114,151,210,219,246,253,262,
296,299,306,376,428,515,666,1310,2611])
assert_almost_equal(trimmed_mean(data,0.1), 343, 0)
assert_almost_equal(trimmed_mean(data,0.2), 283, 0)
def __init__(self, *args, **kwds):
NumpyTestCase.__init__(self, *args, **kwds)
dlist = ['2007-01-%02i' % i for i in range(1,16)]
dates = date_array_fromlist(dlist)
data = masked_array(numeric.arange(15), mask=[1,0,0,0,0]*3)
self.d = (dlist, dates, data)
def __init__(self, *args, **kwds):
NumpyTestCase.__init__(self, *args, **kwds)
dlist = ['2007-01-%02i' % i for i in range(1,16)]
dates = date_array_fromlist(dlist)
data = masked_array(numeric.arange(15), mask=[1,0,0,0,0]*3, dtype=float_)
self.d = (time_series(data, dlist), data, dates)
data['dates']['t'] = ts.Date(freq='t', year=2004, month=1, day=1, hour=0, minute=0)
data['dates']['s'] = ts.Date(freq='s', year=2004, month=1, day=1, hour=0, minute=0, second=0)
for freq in data['dates']:
data['darrays'][freq] = ts.date_array(start_date=data['dates'][freq], length=10)
data['cser']['date_'+freq] = data['darrays'][freq]
data['cser']['bool'] = [True, False, True, False, True, True]
data['cser']['int32'] = np.arange(6).astype(np.int32)
data['cser']['int64'] = np.arange(6).astype(np.int64)
data['cser']['float32'] = np.arange(6).astype(np.float32)
data['cser']['float64'] = np.arange(6).astype(np.float64)
data['cser']['str'] = ["asdf", "aasssssssss", "zzzzzzzzzzzz", "", "blah"]
for x in data['cser']:
data['cser'][x] = ma.masked_array(data['cser'][x])
data['tser'][x] = ts.time_series(data['cser'][x], start_date=data['dates']['a'])
for freq in data['dates']:
data['freqs'][freq] = ts.time_series(np.arange(20).astype(np.float32), start_date=data['dates'][freq])
# test writing for all data types as time series and as case series
for x in data['tser']:
data['tser'][x][1] = ma.masked
data['cser'][x][1] = ma.masked
# series for testing appending data to an existing series
appendTSer = ts.time_series(np.arange(10, 15).astype(np.float32), freq='A', start_date=ts.Date(freq='A', year=2007))
appendCSer = np.arange(10, 15).astype(np.float32)
# series for testing writing over a specified range
def test_trimmedmeanci(self):
"Tests the confidence intervals of the trimmed mean."
data = masked_array([545,555,558,572,575,576,578,580,
594,605,635,651,653,661,666])
assert_almost_equal(trimmed_mean(data,0.2), 596.2, 1)
assert_equal(numpy.round(trimmed_mean_ci(data,0.2),1), [561.8, 630.6])
def test_mjci(self):
"Tests the Marits-Jarrett estimator"
data = masked_array([ 77, 87, 88,114,151,210,219,246,253,262,
296,299,306,376,428,515,666,1310,2611])
assert_almost_equal(mjci(data),[55.76819,45.84028,198.8788],5)
