analysis.R

# Analysis file for measuring segregation
#

# Load libraries
library(dplyr)

# Get functions for segregation metrics
source('metric_functions.R')

# Get geographical data (to add to city data)
geo.data <- read.delim('data/Gaz_tracts_national.txt') %>%
    select(GEOID,
           ALAND) %>%
    rename(area = ALAND)

# store all data as a list of dataframes
# help from: http://stackoverflow.com/questions/5319839/,
#            http://stackoverflow.com/questions/1105659/
folder.location <- file.path('data/prepped')

filenames <- list.files(path = folder.location,
                        pattern = '*.csv')

city.data <- lapply(file.path(folder.location,
                              filenames),
                    read.csv,
                    stringsAsFactors = FALSE) %>%

    # make list names nice
    setNames(gsub('_race.csv',
                  '',
                  filenames,
                  fixed = TRUE)) %>%

    # Remove "-exclude" suffix from some GEOIDs
    # and convert to double
    lapply(function(df) {
        df <- df %>%
            mutate(GEOID = as.double(gsub('-exclude',
                                          '',
                                          GEOID,
                                          fixed = TRUE)))
        return(df)
    }) %>%

    # add geographical data
    lapply(function(df) {
        df <- df %>%
            inner_join(geo.data,
                       by = "GEOID")
        return(df)
    })

# Compute metrics for each city
city.names <- c(
    'Baltimore',
    'Charleston',
    'Chicago',
    'Columbus',
    'Dayton',
    'Denver',
    'Kansas City',
    'Memphis',
    'Milwaukee',
    'Oklahoma City',
    'Pittsburgh',
    'St. Louis',
    'Syracuse',
    'Wichita')

seg.metrics <- as.data.frame(cbind(
    City = city.names,
    Gini = sapply(city.data,
                  giniCoeff),
    Correlation = sapply(city.data,
                         correlationRatio),
    Delta = sapply(city.data,
                   deltaIndex)),
    stringsAsFactors = FALSE) %>%

    mutate(Gini = as.numeric(Gini),
           Correlation = as.numeric(Correlation),
           Delta = as.numeric(Delta))