def sample_clonesize_stat(sample, samdir, freqs=[], numtop=50, args=None):
# calculate: number of clones/ counts that lie within each freq
# range. Note: freqs must be sorted, or the func will sort it
# <numtop>: number of top clones whose freqs will be report
stat = CloneSizeStat(sorted(freqs))
stat.set_sample_info(sample)
if args:
numtop = args[0]
clones = libsample.sample_all_clones(samdir)
sorted_clones = sorted(clones, reverse=True, key=lambda c: c.freq)
for index, clone in enumerate(sorted_clones):
if index < numtop:
stat.topfreqs.append(clone.freq)
for i, minfreq in enumerate(stat.freqs):
maxfreq = float('inf')
if i + 1 < len(stat.freqs):
maxfreq = stat.freqs[i + 1]
if minfreq <= clone.freq and clone.freq < maxfreq:
stat.numclones[i] += 1
stat.counts[i] += clone.count
# convert to frequencies:
stat.numclones = [libcommon.get_pc(c, stat.numclone) for c in
stat.numclones]
stat.counts = [libcommon.get_pc(c, stat.size) for c in stat.counts]
# get cumulative stats:
stat.numclones_cumul = libcommon.get_cumulative(stat.numclones)
stat.counts_cumul = libcommon.get_cumulative(stat.counts)
stat.topfreqs_cumul = libcommon.get_cumulative(stat.topfreqs, True)
return stat
def test_get_cumulative(self):
vals = [1, 0, -1, 0.5, 2.5]
forward_vals = [1, 1, 0, 0.5, 3]
cumu_vals = lcommon.get_cumulative(vals, forward=True)
self.assertEqual(forward_vals, cumu_vals)
backward_vals = [3, 2, 2, 3, 2.5]
cumu_vals = lcommon.get_cumulative(vals)
self.assertEqual(backward_vals, cumu_vals)
