def reduce(self, resultset, series, data=None):
key = series['data']
smeta = resultset.meta('SERIES_META').get(key)
if smeta and 'OWD_UP_MEAN' in smeta and not self.cutoff:
delay = smeta['OWD_UP_MEAN'] + 2 * smeta['IPDV_UP_MEAN']
lossrate = smeta['PACKET_LOSS_RATE']
else:
delay, lossrate = self._calc_delay_loss(resultset, key)
mos = mos_score(delay, lossrate)
return mos
def reduce(self, resultset, series, data=None):
key = series['data']
data = self._get_series(resultset, key)
if not data:
return None
jitter_samples = []
delay_samples = []
loss = 0
last_delay = -1
last_seq = -1
for d in data:
if last_seq > -1 and d['seq'] - last_seq > 1:
loss += 1
last_seq = d['seq']
if last_delay > -1:
jitter_samples.append(abs(last_delay - d['val']))
last_delay = d['val']
delay_samples.append(d['val'])
delay = numpy.mean(delay_samples) + 2 * numpy.mean(jitter_samples)
lossrate = loss / len(data)
mos = mos_score(delay, lossrate)
return mos
def reduce(self, resultset, series, data=None):
key = series['data']
data = self._get_series(resultset, key)
if not data:
return None
jitter_samples = []
delay_samples = []
loss = 0
last_delay = -1
last_seq = -1
for d in data:
if last_seq > -1 and d['seq'] - last_seq > 1:
loss += 1
last_seq = d['seq']
if last_delay > -1:
jitter_samples.append(abs(last_delay - d['val']))
last_delay = d['val']
delay_samples.append(d['val'])
delay = np.mean(delay_samples) + 2 * np.mean(jitter_samples)
lossrate = loss / len(data)
mos = mos_score(delay, lossrate)
return mos