The reason was mainly because, everytime I pay my rent in SF, my heart bleeds a little. People always say that San Francisco was one of the most expensive places to live in the United States, I want to know how true is that.
The data used in the visualization above is from the zillow research website. The specific data set used is the Neighborhood category, ZRI Time Series: Multifamily, SFR, Condo/Co-op ($) which takes into account of all home and apartment types.
The original data for prices are based on the Zillow Rent Index(ZRI). According to the Zillow Research website, the are 3 main steps involved in constructing the ZRI—Calculate Raw Median Rent Zestimates, Apply Simple 3-Month Moving Average, and Final Quality Control.
The ZRI is calculated using Zillow's proprietary statistical and machine learning models. Within each county or state, the models observe recent rental listings and learn the relative contribution of various home attributes in predicting prevailing rents. These home attributes include physical facts about the home, prior sale transactions, tax assessment information and geographic location as well as the estimated market value of the home. For more info, go to the Zillow page on methodology.
In it's original form, the data was in wide format, instead of the usual time-series long format. R was used to reshape the data to long format for easier visualizing. The data was also further processed by filtering out only continental US data and sorting was done for convenience.
Because the Zillow Website did not have a dataset aggregated on City Price and due to licensing terms that requires the data to be in a derivative form, the neighborhood data was aggregated by simply taking the average on the ZRI prices.
The data was divided into 3 files—state,county, and city—by aggregating neighborhood prices to the specified categories. The R-package "ggmap" was also used to iteratively download latitude and logitude data for each city from the google map api.
Click to view R Script used.
It was interesting to see that Delaware had the highest rent out of all the states.
Abnormally high median rent in Collier. After some investigation, the reason for the high rent was found. It was due to the nice location being by the beach and the fact that many rental houses listed in that area there cost #20,000 and more.
Suprised to see SF wasn't even in the top 10. Naples was very high for the same reason as the Collier county.
In conclusion, SF is not as expensive as I thought in terms of rent. However, bear in my that the data includes prices for all housing types whether 1 bed room or more, plus the fact that the aggregate data is based on the median, this visualization would not be as helpful for people looking for smaller or specific housing across the states and counties. This visualization is however, great if you wan to know which areas had the greater porportions of high rent housing.
