driving-dist-analysis.Rmd

---
title: "Driving Distances to the Closest Cites Across the U.S."
author: "Jeff Erickson"
date: "Saturday, September 13, 2014"
output: html_document
---

### LOAD LIBRARIES

```{r}
library(ggplot2)
library(scales)
library(knitr)
library(rgeos)
library(rgdal)
```

### HELPER FUNCTIONS

```{r}
z.score <- function(x) { #Calc z-scores
  #arg: a vector of values to z-score
  return((x - mean(x)) / sd(x))
}
```

### IMPORT AND SUBSET

```{r}
driving.dists <- read.csv("datasets/county-city-driving-dist.csv", header=TRUE, as.is=TRUE)

driving.dists.shortest <- driving.dists[which(driving.dists$shortest_flag == 1), ]
```

### CALCULATE Z-SCORE

```{r}
#Z-Score of driving distance
driving.dists.shortest$z.score <- z.score(driving.dists.shortest$driving_distance)
```

### BASIC INVESTIGATIONS

```{r}
summary(driving.dists.shortest$driving_distance)

ggplot(driving.dists.shortest, aes(x=driving_distance)) +
  geom_density()

ggplot(driving.dists.shortest, aes(x=z.score)) +
  geom_density()
```

Top 10 Counties from Major U.S. Cities:
```{r}
kable(head(driving.dists.shortest[order(-driving.dists.shortest$driving_distance),], n=10), format="markdown", row.names=FALSE)
```

### HEAT MAP

Prepare the shapes for mapping.
```{r}
county.shapes <- readOGR(dsn="datasets", "gz_2010_us_050_00_500k")
county.shapes.f <- fortify(county.shapes, region="GEO_ID")

county.shapes.f <- merge(county.shapes.f, driving.dists.shortest, by.x="id", by.y="fips")
```

...and map!
```{r} 
ggplot(county.shapes.f[order(county.shapes.f$order), ], aes(x=long, y=lat, group=group)) +
  geom_polygon(colour="black", aes(fill=z.score)) +
  scale_fill_gradientn("Driving Distance\n(z-Score)", colours=c("green", "white", "red"), values=c(-5, 0, 5), rescaler=function(x, ...) x, oob=identity) +
  coord_equal() +
  ggtitle(expression(atop("Driving Distance to Nearest Large U.S. City", atop("Compared to Average", "")))) +
  theme(axis.title=element_blank(), axis.text=element_blank(), axis.ticks=element_blank(), panel.grid=element_blank())
```