Group Members: Adrian, Tristan, Amy Tran, Masa
build 5 models-Describe the car model you selected, the variables (including price, age, location, and mileage) you chose to look at, etc.
The goal of this project is to fit models of used car prices with variables our group deemed important that may influence the used car prices. Our group used the website www.cars.com to obtain our data of used Honda Civic cars for this project. Used Honda civic car samples were taken from Los Angeles, California (zip code 90053) and Boston, Massachusetts (zip code 02118). We decided that because we are looking up data on used cars and putting them into a spreadsheet by hand, 30 cases at each location would be feasible and manageable. The search setting on www.cars.com was set to find Honda Civics within 10 miles of each location. Beside that, there were no settings for a specific price, year, mileage, or model, because we want to have a representative sample. We tried to randomly chose cases for each location to get a representative sample. We picked different cases from all the pages that came up when we hit search without repeating cases. We took note on the price, mileage, year, and model of the Honda Civics.
The models we fitted are price of used Honda civics to mileage of the cars, LIST OTHER MODELS HERE
dataSource = "https://docs.google.com/spreadsheet/pub?key=0AsXterp7HII8dE9LbGkzTlNBQlprZWVydFZiLTRRREE&output=csv"
cars = fetchGoogle(dataSource)
xyplot(Price~Mileage, data=cars,
ylab="Price ($)", xlab="Miles")
The scatter plot of Mileage, in miles, versus the price of the used Honda Civic car, in dollars, shows that the price of the Honda Civic decreases with increasing mileage.
densityplot(~Mileage, data=cars)
The density plot for the price of used Honda Civic cars displays a fairly normal bell-curve distribution beside the shape to the right of the mean. To the right of the mean, the density decrease like a normal distribution but then slightly increases before decreasing again. An explanation for this is that some Honda Civics sampled with higher mileages are actually priced higher than other samples of Honda Civics with lower mileages. This is reflective of the above scatter plot of Mileage versus price. From the scatter plot at about $10,000, there seems to be a wide range of mileage of about 40,000 miles to 115,000 miles.
mod1 = lm( Price ~ Mileage, data=cars)
coef(mod1)
## (Intercept) Mileage
## 18129.0282 -0.1019
xyplot(fitted(mod1)+Price~Mileage,data=cars)
For every 1 mile increase in mileage of a used Honda Civic car, the price of the car is cheaper by 0.1019 cents or about 11 cents. That is, there is decrease in price of a used Honda Civic of about $0.11 per mile put onto that car. The average price of a used Honda Civic car with 0 miles on the car costs $18,129.0282.