def test_simple_peak_find():
trace = generate_gaussian()
peaks = simple_peak_find(trace,
init_slope=0.2,
start_slope=0.2,
end_slope=0.2,
min_peak_height=0.01)
# TODO: better asserts
assert len(peaks) > 0
def test_simple_peak_find():
trace = generate_gaussian()
peaks = simple_peak_find(trace, init_slope=0.2, start_slope=0.2,
end_slope=0.2, min_peak_height=0.01)
# TODO: better asserts
assert len(peaks) > 0
def integrate(s, peak_list=None, baseline=None, overlap='drop'):
"""
"""
# for convenience, if no peak_list is provided create our own
if peak_list is None:
peak_list = simple_peak_find(s)
peaks = []
for _, pks in _get_windows(peak_list):
temp_pks = []
pks = sorted(pks, key=lambda p: p[0])
if 'y0' in pks[0][2] and 'y1' in pks[-1][2]:
y0, y1 = pks[0][2]['y0'], pks[-1][2]['y1']
else:
y0 = s.get_point('!', pks[0][0])
y1 = s.get_point('!', pks[-1][1])
ys = np.array([y0, y1])
xs = np.array([pks[0][0], pks[-1][1]])
# go through list of peaks to make sure there's no overlap
for t0, t1, hints in pks:
# figure out the y values (using a linear baseline)
hints['y0'] = np.interp(t0, xs, ys)
hints['y1'] = np.interp(t1, xs, ys)
# if this peak totally overlaps with an existing one, don't add
if sum(1 for p in temp_pks if t1 <= p[1]) > 0:
continue
overlap_pks = [p for p in temp_pks if t0 <= p[1]]
if len(overlap_pks) > 0:
# find the last of the overlapping peaks
overlap_pk = max(overlap_pks, key=lambda p: p[0])
# get the section of trace and find the lowest point
over_ts = s.trace('!', twin=(t0, overlap_pk[1]))
min_t = over_ts.times[over_ts.y.argmin()]
# delete the existing overlaping peak
for i, p in enumerate(temp_pks):
if p == overlap_pk:
del temp_pks[i]
break
# interpolate a new y value
y_val = np.interp(min_t, xs, ys)
overlap_pk[2]['y1'] = y_val
hints['y0'] = y_val
# add the old and new peak in
temp_pks.append((overlap_pk[0], min_t, overlap_pk[2]))
temp_pks.append((min_t, t1, hints))
else:
temp_pks.append((t0, t1, hints))
# none of our peaks should overlap, so we can just use
# simple_integrate now
peaks += simple_integrate(s, temp_pks)
for p in peaks:
p.info['p-create'] = p.info['p-create'].split(',')[0] + \
',drop_integrate'
return peaks