All steps to clean / transform the data must be documented, with the full source code and/or a detailed log of what was done as appropriate, and the resulting database must be saved as separate files.

Fields in policies.csv

Field label	Description	Type of data	Range
policy_id	Unique customer identifier	ID	policy_##### (may be shorter than 5 digits)
Quote_dt	Date the quote was submitted	Approx. continuous	2015 - 2018 (yyyy-mm-dd)
quoted_amt	Quote amount (US dollars)	Approx. continuous (positive)	$15 - $108,608 or NA
Prior_carrier_grp	Prior carrier group	Factor	Carrier_1 to 8, Other, NA
Cov_package_type	Level of coverage needed	Factor	Low, Medium, High, NA
discount	Whether or not a discount was applied to the quote amount	Boolean	Yes, No
number_drivers	Number of drivers	Discrete (positive)	1 - 6
credit_score	Credit score of primary policy holder	Approx. continuous (positive)	369 - 850, NA (credit score can go as low as 350)
num_loaned_veh	Number of vehicles on policy that have a loan associated with them	Discrete (nonnegative)	0 - 3 Inaccurate; do not use
num_owned_veh	Number of owned vehicles on the policy	Discrete (nonnegative)	1 - 3 Inaccurate; do not use
num_leased_veh	Number of leased vehicles on the policy	Discrete (nonnegative)	0 - 2 Inaccurate; do not use
total_number_veh	Total number vehicles on the policy	Discrete (nonnegative)	1 - 8 Inaccurate; do not use
primary_parking	Where car(s) are primarily parked	Factor	home/driveway, parking garage, street, unknown
CAT_zone	Catastrophe risk zone	Factor	1 - 5, NA
Home_policy_ind	Does customer has existing home insurance policy with Peace of Mind	Boolean	Y, N
zip	US zip code of policy holder	Factor	#####, NA
state_id	State of policy holder	Factor	AL, CT, FL, GA, MN, NJ, NY, WI
county_name	County of policy holder	Factor	Too numerous to list
Agent_cd	Unique agent code (8 digits)	ID	########, NA
split	Train/Test split	Factor	Test, Train
convert_ind	Conversion indicator (0=no, 1=yes). This is the response variable	Boolean	0, 1 for train data; NA for test data

Fields in drivers.csv

Field label	Description	Type of data	Range
policy_id	Unique customer identifier	ID	policy_##### (may be shorter than 5 digits)
gender	Gender of driver	Factor / Boolean	F, M
age	Age of driver	Approx. continuous (nonnegative)	16 - 147
high_education_ind	Higher education indicator	Boolean	0, 1, NA
safty_rating	Safety rating index of driver	Approx. continuous	-28 - 100, NA
living_status	Driver's living status	Factor	dependent, own, rent, NA

Fields in vehicles.csv

Field label	Description	Type of data	Range
policy_id	Unique customer identifier	ID	policy_##### (may be shorter than 5 digits)
car_no	Unique car identifier (per policy)	ID	1 - 8
ownership_type	Whether the car is loaned, owned or leased	Factor	leased, loaned, owned
color	Vehicle color	Factor	black, blue, gray, red, silver, white, other, NA
age	Vehicle age	Discrete (nonnegative)	0 - 16, NA
make_model	Make and model of the vehicle	Factor	Too numerous to list