Analyzing points in a polygon in R

points in a polygon

# Calculating points in a polygon

by_tract <- over(sp, towntracts_only)

# What just happened: Every point in the list now has a corresponding census tract 
kable(head(by_tract, 5))

# Use dplyr to gather up and count how many instances of census tracts there are

require(dplyr)

by_tract <- by_tract %>%
  group_by(GEOID10) %>%
  summarise(total=n())

# Get rid of the census tracts with no data
by_tract <- by_tract[!is.na(by_tract$GEOID10),]

kable(head(by_tract,5))

# Rename the columns of this datframe so it can be joined to future data
colnames(by_tract) <- c("id", "total")

# Changing the GEOID number to character so it can be joined to future data
by_tract$id <- as.character(by_tract$id)

# Bring in a dataframe that has matches census tract ID numbers to town names
tracts2towns <- read.csv("data/tracts_to_towns.csv", stringsAsFactors=FALSE)

kable(head(tracts2towns, 5))

# Changing the column names so it can be joined to the by_tract dataframe
colnames(tracts2towns) <- c("id", "town_name")

# Changing the GEOID number to character so it can be joined to the by_tract dataframe
tracts2towns$id <- as.character(tracts2towns$id)

# Adding a 0 to the front of the GEOID string because it was originally left out when it was imported
tracts2towns$id <- paste0("0", tracts2towns$id)

# Bringing in a library to deal with strings
library(stringr)

# Eliminating leading and trailing white space just in case
tracts2towns$town_name <- str_trim(tracts2towns$town_name)

# Joining the by_tract dataframe to the tracts2towns dataframe

by_tract <- left_join(by_tract, tracts2towns)