retention analysis by cohort

retention_analysis.py

import csv


org_data = open("data2.csv")

data_alpha, data_beta, org_first_seen = list(), list(), list() 

with open('data2.csv', 'rb') as csvfile_A:
	csvfile_A = csv.reader(csvfile_A)
	for item in csvfile_A:
		# print(item)
		data_alpha.append(item)

def parse_data(data):
	parsed_data = list()
	# Get the ordered date and Org ID, and Order status
	parsed_data = [data[1], data[3], data[4]]
	# Convert ordered date to yyyy-mm
	parsed_data[0] = parsed_data[0][1:8]
	# convert order status to lowercase and delete extra space
	parsed_data[1] = parsed_data[1][1:].lower()
	return parsed_data

def generate_clean_data(raw_data, clean_data):
	clean_data = list()
	for item in raw_data:
		clean_data.append(parse_data(item))
	return clean_data

def add_to_cohort(data, existing_cohort, cohort_date):
	order_date = data[0]
	order_status = data[1]
	org_id = data[2]
	if order_date == cohort_date and order_status == 'completed' and org_id not in existing_cohort:
		return True
	else:
		return False 


data_beta = generate_clean_data(data_alpha, data_beta)


def generate_cohort(data_beta, cohort, date):
	for item in data_beta:
		if add_to_cohort(item, cohort, date) == True:
			cohort.append(item[2])

april_cohort, may_cohort, june_cohort = list(), list(), list()

generate_cohort(data_beta, april_cohort, '2018-04')
generate_cohort(data_beta, may_cohort, '2018-05')
generate_cohort(data_beta, june_cohort, '2018-06')

print(len(april_cohort))
print(len(may_cohort))
print(len(june_cohort))

returning_orgs_april_may = 0
for item in april_cohort:
	if item in may_cohort:
		# print(item)
		returning_orgs_april_may += 1
		# print(returning_orgs_april_may, '!!!')

print('Retention is')
print(float(returning_orgs_april_may)) / float(len(april_cohort))