def addtrans(raw, y, label, plot):
"""
Add little triangles marking where the transits are.
"""
if raw.dtype.names.count('finj') == 1:
finjkw = dict(color='m', mew=2, marker=7, ms=5, lw=0, mfc='none')
if label == True:
finjkw['label'] = 'Injected Transits'
ym = ma.masked_array(y, raw['finj'] > -1e-6)
t = raw['t']
sL = ma.clump_unmasked(ym)
id = [s.start for s in ma.clump_unmasked(ym)]
plot(t[id], ym[id] + 0.001, **finjkw)
def adapt_mask_to_roi(consmask, roifp, chrom):
"""
:param consmask:
:param roifp:
:param chrom:
:return:
"""
roimask = np.ones(consmask.size, dtype=np.bool)
opn, mode = text_file_mode(roifp)
chrom_found = False
with opn(roifp, mode) as rois:
for line in rois:
if not line or line.startswith('#'):
continue
cols = line.strip().split()
if cols[0] != chrom:
continue
chrom_found = True
roimask[int(cols[1]):int(cols[2])] = 0
comb_mask = consmask | roimask # 0 | 0 = 0
assert not comb_mask.sum() == consmask.size,\
'All positions on chromosome {} masked - ' \
'any regions for that chromosome in ROI file: {}'.format(chrom, chrom_found)
pos = np.arange(consmask.size)
pos = msk.array(pos, mask=comb_mask)
select = np.array([k for k, g in itt.groupby(~comb_mask)], dtype=np.bool)
splits = np.array(list(
itt.chain.from_iterable(
(item.start, item.stop) for item in msk.clump_unmasked(pos))),
dtype=np.int32)
return splits, select
def wet_spells(rainfall, threshold=0, start=True, mode='both'):
"""
Compute the wet spells for a series of precipitations
Parameters
----------
rainfall : TimeSeries
TimeSeries of precipitations.
threshold : float, optional
Minimum amount of precipitation defining a wet day.
start : boolean, optional
Whether the spells are associated with the first or last day of the spell.
mode : {'durations', 'intensities', 'all'}
Whether to return only durations, intensities or both.
Returns
-------
wet_spells : TimeSeries
A :class:`TimeSeries` giving the duration and/or intensity of a spell
at either the first or last date.
"""
rdates = getattr(rainfall, 'dates', None)
rdata = ma.masked_array(rainfall, subok=False)
condition = ma.masked_less(rdata, threshold)
slices = ma.clump_unmasked(condition)
# Get the durations and starting dates of each spell
mode = (mode or 'both').lower()[0]
if mode == 'd':
# Durations
result = np.array([s.stop - s.start for s in slices], dtype=int)
elif mode == 'i':
# Intensities
result = np.array([rdata[s].sum() for s in slices])
else:
durations = [s.stop - s.start for s in slices]
intensities = [rdata[s].sum() for s in slices]
result = np.array(zip(durations, intensities),
dtype=[('durations', int), ('intensities', float)],)
if rdates is None:
return result
if start:
dates = rdates[[s.start for s in slices]]
else:
dates = rdates[[s.stop - 1 for s in slices]]
ensoi = getattr(rainfall, 'ensoindicator', None)
if mode in 'id':
spells = enso.climate_series(result, dates=dates, ensoindicator=ensoi)
else:
spells = enso.climate_records(result, dates=dates, ensoindicator=ensoi)
return spells
def process_alignments(parameter_set):
chrom, chrom_size, mapq_low, mapq_high, quantifier, assembly_threshold, dump_type, dump_counts, aln_store, aln_keys = parameter_set
if quantifier in ['all', 'any']:
bool_mask = build_boolean_mask(
chrom,
chrom_size,
mapq_low,
mapq_high,
quantifier,
aln_store,
aln_keys
)
if dump_type == 'alignments':
# the mask is build by iterating over / masking
# alignment regions, so need to _invert_
# the mask to mask out all unaligned
# (unselected) regions
genomic_coordinates = ma.masked_array(np.arange(chrom_size, dtype=np.int32), mask=~bool_mask)
elif dump_type == 'inverse':
# mask can be used as-is, all unaligned
# (unselected) regions will be dumped
genomic_coordinates = ma.masked_array(np.arange(chrom_size, dtype=np.int32), mask=bool_mask)
else:
raise ValueError('Unexpected value for "dump_type": {}'.format(dump_type))
elif quantifier in ['lt', 'geq']:
quant_mask = build_quantitative_mask(
chrom,
chrom_size,
mapq_low,
mapq_high,
quantifier,
assembly_threshold,
dump_counts,
aln_store,
aln_keys
)
# negating mask: see reason above
genomic_coordinates = ma.masked_array(np.arange(chrom_size, dtype=np.int32), mask=~quant_mask)
else:
raise ValueError('Unsupported quantifier: {}'.format(quantifier))
regions = [(s.start, s.stop) for s in ma.clump_unmasked(genomic_coordinates)]
return chrom, regions
def dry_spells(rainfall, threshold=0.254, start=True):
"""
Compute the dry spells for a series of precipitations
Parameters
----------
rainfall : TimeSeries
TimeSeries of precipitations.
threshold : float, optional
Minimum amount of precipitation defining a wet day.
start : boolean, optional
Whether the spells are associated with the first or last day of the spell.
Returns
-------
dry_spells : TimeSeries
A :class:`TimeSeries` giving the duration of the spell at either the first
or last date.
"""
rdates = getattr(rainfall, 'dates', None)
rdata = ma.masked_array(rainfall, subok=False)
condition = ma.masked_greater(rdata, threshold)
slices = ma.clump_unmasked(condition)
# Get the durations and starting dates of each spell
durations = [s.stop - s.start for s in slices]
# Give up if we have no dates
if rdates is None:
return np.array(durations)
# Get the dates, then...
if start:
dates = rdates[[s.start for s in slices]]
else:
dates = rdates[[s.stop - 1 for s in slices]]
ensoi = getattr(rainfall, 'ensoindicator', None)
spells = enso.climate_series(durations, dates=dates, ensoindicator=ensoi)
return spells
