Gold Analysis .txt

import types 
import sqlalchemy as sqlalchemy
from dateutil.relativedelta import relativedelta

def dbg(dbobj, flat=False):
	object_list = [dbobj] if (type(dbobj) is not types.ListType) and (type(dbobj) is not sqlalchemy.orm.collections.InstrumentedList) else dbobj
	attrs = engine.execute("select * from " + object_list[0].__tablename__).keys()
	for obj in object_list: 
		if not flat:
			pprint([(attr, getattr(obj, attr)) for attr in attrs])
		else:
			print([(attr, getattr(obj, attr)) for attr in attrs])


#Gold Churn 

start_date = datetime.date(2013, 1, 1)
totango_start = datetime.date(2013, 11, 1)



accounts = s.query(Summary).filter(Summary.plan == 'gold', Summary.start_date >= start, Summary.churn==True).all()
date_range = [(start_date + relativedelta(months=i)) for i in range(0,13)]

# months active == 1 per month 
[len([a for a in accounts if a.start_date.month==d.month and a.start_date.year==d.year and a.months_active==1]) for d in date_range]
#2/18 
[19, 10, 28, 19, 19, 30, 26, 25, 31, 53, 53, 31, 12]

accounts = s.query(Summary).filter(Summary.plan == 'gold', Summary.start_date >= start).all()
# start per month 
[len([a for a in accounts if a.start_date.month==d.month and a.start_date.year==d.year]) for d in date_range]
# 2 / 18 
[93, 73, 83, 102, 103, 114, 123, 116, 102, 134, 142, 170, 121]






############ TOTANGO ANALYSIS ###################
# months active over time 
# How do we calculate months active? Is it a delta in days? 


accounts = s.query(Summary).filter(Summary.plan == 'gold', Summary.start_date >= totango_start, Summary.churn==True).all()
date_range = [(totango_start + relativedelta(months=i)) for i in range(0,3)]

def getAttr(attr, a):
	return sum([getattr(m, attr) or 0 for m in a.totango_months]) / a.months_active

# months_active == 1 


# avg view results for accounts with months active == 1 and who churned 
m = 1
sum([getAttr("view_results", a) for a in accounts if a.months_active==m]) / float(len(filter(lambda x: x.months_active==m, accounts)))
m = 2
sum([getAttr("view_results", a) for a in accounts if a.months_active==m]) / float(len(	(lambda x: x.months_active==m, accounts)))

accounts = s.query(Summary).filter(Summary.plan == 'gold', Summary.start_date >= totango_start, Summary.churn==False).all()

# avg view results per month 
[sum([getAttr("view_results", a) for a in accounts if a.months_active==1 and a.start_date.month==d.month and a.start_date.year==d.year]) / len([getAttr("view_results", a.totango_months) for a in accounts if a.months_active==1 and a.start_date.month==d.month and a.start_date.year==d.year]) for d in date_range]


 
 ###plat###
accounts = s.query(Summary).filter(Summary.plan == 'gold', Summary.start_date >= totango_start, Summary.churn==True).all()

debug = s.query(Summary).filter(Summary.plan == 'gold', Summary.start_date >= totango_start, Summary.churn==True).all()