def median_entity_salary(self):
q = Salary.objects.filter(
Q(job__position__employer__parent=self.object)
| Q(job__position__employer=self.object))
results = q.all().aggregate(
median=Percentile('amount', 0.5, output_field=FloatField()))
return results['median']
def get_queryset(self):
req=self.request
airport1 = req.GET.get('airport1')
airport2 = req.GET.get('airport2')
carrier = req.GET.get('carrier')
if airport1 and airport2 and carrier:
return Data.objects.all().filter(
Q(airport_code=airport1) |
Q(airport_code=airport2)
).filter(
carrier_code=carrier
).values(
'statistics_minutesdelayed_carrier',
'statistics_minutesdelayed_lateaircraft'
).annotate(
mean_carier = Avg('statistics_minutesdelayed_carrier'),
mean_lateaircraft = Avg('statistics_minutesdelayed_lateaircraft'),
median_carier=Percentile('statistics_minutesdelayed_carrier', 0.5, output_field=models.FloatField()),
median_lateaircraft=Percentile('statistics_minutesdelayed_carrier', 0.5, output_field=models.FloatField()),
std_dev_carrier = StdDev('statistics_minutesdelayed_carrier'),
std_dev_lateaircraft = StdDev('statistics_minutesdelayed_lateaircraft')
)
elif airport1 and airport2:
return Data.objects.all().filter(
Q(airport_code=airport1) |
Q(airport_code=airport2)
).values(
'statistics_minutesdelayed_carrier',
'statistics_minutesdelayed_lateaircraft'
).annotate(
mean_carier = Avg('statistics_minutesdelayed_carrier'),
mean_lateaircraft = Avg('statistics_minutesdelayed_lateaircraft'),
median_carier=Percentile('statistics_minutesdelayed_carrier', 0.5, output_field=models.FloatField()),
median_lateaircraft=Percentile('statistics_minutesdelayed_carrier', 0.5, output_field=models.FloatField()),
std_dev_carrier = StdDev('statistics_minutesdelayed_carrier'),
std_dev_lateaircraft = StdDev('statistics_minutesdelayed_lateaircraft')
)
def employer_median_salaries(self):
if not hasattr(self, '_employer_median_salaries'):
median_base_pay = Percentile('positions__jobs__salaries__amount',
0.5,
filter=self.salary_q,
output_field=FloatField())
median_extra_pay = Percentile(
'positions__jobs__salaries__extra_pay',
0.5,
filter=self.salary_q,
output_field=FloatField())
median_total_pay = Percentile(NullIf(
Coalesce('positions__jobs__salaries__amount', 0) +
Coalesce('positions__jobs__salaries__extra_pay', 0), 0),
0.5,
filter=self.salary_q,
output_field=FloatField())
self._employer_median_salaries = self.employer_queryset.aggregate(
median_base_pay=Coalesce(median_base_pay, 0),
median_extra_pay=Coalesce(median_extra_pay, 0),
median_total_pay=Coalesce(median_total_pay, 0))
return self._employer_median_salaries
def label_timing(request, project_pk):
"""This function finds and returns the requested label time metrics. This is used by
the graphs on the admin page to show how long each labeler is taking.
Args:
request: The POST request
project_pk: Primary key of the project
Returns:
a dictionary of label timing information.
"""
project = Project.objects.get(pk=project_pk)
users = []
users.append(project.creator)
users.extend(
[perm.profile for perm in project.projectpermissions_set.all()])
dataset = []
yDomain = 0
for u in users:
result = DataLabel.objects.filter(
data__project=project_pk,
profile=u).aggregate(quartiles=Percentile(
"time_to_label",
[0.05, 0.25, 0.5, 0.75, 0.95],
continuous=False,
output_field=ArrayField(FloatField()),
))
if result["quartiles"]:
if result["quartiles"][4] > yDomain:
yDomain = result["quartiles"][4] + 10
temp = {
"label": u.__str__(),
"values": {
"Q1": result["quartiles"][1],
"Q2": result["quartiles"][2],
"Q3": result["quartiles"][3],
"whisker_low": result["quartiles"][0],
"whisker_high": result["quartiles"][4],
},
}
dataset.append(temp)
return Response({"data": dataset, "yDomain": yDomain})
def benefit_aggregates(self):
data = self._cache.get('benefit_aggregates', None)
if data is None:
aggregates = Benefit.objects\
.values('fund__name', 'data_year')\
.annotate(median=Percentile('amount', 0.5, output_field=FloatField()), count=Count('id'))
data = {year: {} for year in self.data_years}
for year in data.keys():
agg = [a for a in aggregates if a['data_year'] == year]
for a in agg:
data[year][a['fund__name']] = {
'median': self._format_large_number(a['median']),
'count': self._format_large_number(a['count']),
}
self._cache['benefit_aggregates'] = data
return data
def NearMax(field):
"DB Statistical function for almost the maximum but not quite."
return Percentile(field, 0.95, output_field=FloatField())