Lab 02 - Identifying Groups

Data

The following report identifies specific tax parcels in Downtown Syracuse, New York (USA).

View the dataset documentation here: Syracuse, NY Tax Parcel Data.

Required Packages

The following packages are required for this report and loaded with library().

library(sp)
library(rgdal)
library(geojsonio)                            # Load required libraries

Import Spatial Polygons

We read in spatial polygons ("sp") with function geojson_read() to map tax parcel shapes in Syracuse’s Downtown area.

url <- paste0("https://raw.githubusercontent.com/DS4PS/Data",
              "-Science-Class/master/DATA/downtown-syr.geojson")

downtown <- geojson_read(url, what = "sp")    # Load parcel data

plot(downtown,  
     border = "gray60",
     col = "gray80")                          # Visualize parcels

Example Solutions

The following examples demonstrate how to use conditions to identify specific parcels.

Example: Tax Parcel Acres

Parcels with more than one acre are determined with variable acres and condition > 1.

• Each value in variable acres is tested as greater than one, or > 1
• Values in acres greater than one become TRUE, otherwise FALSE
• All TRUE and FALSE values are stored in object these

result <- downtown$acres > 1                  # Each value in 'acres' is tested
summary(result)

##    Mode   FALSE    TRUE 
## logical     330      59

All TRUE and FALSE values, for each parcel, are converted into colors.

group.colors <- ifelse(test = result,         # 'test =' accepts TRUE or FALSE values
                       yes = "firebrick",     # 'yes =' converts values that are TRUE
                       no = "gray80")         # 'no =' converts values that are FALSE

All tax parcels in downtown are mapped with plot(). Each parcel has one of two colors per the above code.

plot(downtown,                                # Plots object 'downtown'
     border = NA,
     col = group.colors)                      # Red when TRUE; gray when FALSE

Example: Parcels with Single Families

Use variable landuse to determine how a tax parcel is used. Determine all possible values with unique().

unique(downtown$landuse)                      # All unique values in 'landuse'

##  [1] "Parking"            "Commercial"         "Parks"             
##  [4] "Community Services" "Vacant Land"        "Utilities"         
##  [7] "Apartment"          "Recreation"         "Schools"           
## [10] "Religious"          "Industrial"         "Single Family"

One of the values in landuse is “Single Family”. Therefore, the test is whether or not landuse is exactly equal to “Single Family”.

result <- downtown$landuse == "Single Family" # Test if 'landuse' is "Single Family"

group.colors <- ifelse(test = result,         # Provide TRUE or FALSE test results
                       yes = "firebrick",     # If TRUE, make "firebrick"
                       no = "gray80")         # If FALSE, make "gray80"

plot(downtown,
     border = NA, 
     col = group.colors)                      # Plot with respective colors

Questions & Solutions

The following questions ask you to map your results, write your answer, or both. The first question has been completed for you.

Question 1: Vacant Lots

Question: Where are the majority of vacant lots located in the downtown? Map your results.

Note: This solution has been provided for you.

unique(downtown$landuse)                      # Print unique values in 'landuse'

##  [1] "Parking"            "Commercial"         "Parks"             
##  [4] "Community Services" "Vacant Land"        "Utilities"         
##  [7] "Apartment"          "Recreation"         "Schools"           
## [10] "Religious"          "Industrial"         "Single Family"

Now, we create our test statement, == "Vacant Land", and store the results in result.

result <- downtown$landuse == "Vacant Land"

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Question 2: Parking Lots

Question: How many parking lots are in downtown Syracuse? Map your results.

Answer: There are 78 parking lots in downtown Syracuse.

# Use function: 'sum()'
#unique(downtown$landuse)
# Use variable: 'landuse'
sum(downtown$landuse=="Parking")

## [1] 78

# Assign test (logical) output to object 'result' by replacing 'FALSE'

result <- downtown$landuse=="Parking"

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Question 3: New Construction

Question: Where is new construction located in the city?. Map your results.

Note: You may use >= (greater than or equal) or > (greater than) 1980.

# Use variable: 'yearbuilt'
#is.data.frame(downtown)
#summary(downtown)
# Find the number of buildings constructed either after or during and after than 1980
#summary(downtown$yearbuilt)
#is.numeric(downtown$yearbuilt)
sum(downtown$yearbuilt >= 1980, na.rm = TRUE)

## [1] 46

# Assign test (logical) output to object 'result' by replacing 'FALSE'

result <- downtown$yearbuilt >= 1980

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Question: What proportion of commercial properties are built since 1980?

Answer: 14.35% of commercial properties were built since 1980.

unique(downtown$landuse)

##  [1] "Parking"            "Commercial"         "Parks"             
##  [4] "Community Services" "Vacant Land"        "Utilities"         
##  [7] "Apartment"          "Recreation"         "Schools"           
## [10] "Religious"          "Industrial"         "Single Family"

# Use variable: 'landuse'
total <- sum(downtown$landuse == "Commercial", na.rm=TRUE)
# Use variable: 'yearbuilt'
# Count only commercial properties with buildings constructed during or later than 1980
prop <- sum(downtown$landuse == "Commercial" & downtown$yearbuilt >= 1980, na.rm=TRUE)
prop/total*100

## [1] 14.35407

#visual representation
result <- (downtown$landuse == "Commercial" & downtown$yearbuilt >= 1980)

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Question 4: Parcels Valued Over $10M

Question: How many parcels have assessed values over $10 million? Map your results.

Answer: 16 parcels have assessed values over $10 M.

# Use function: 'sum()'
# Use variable: 'assessedval'
#summary(downtown$assessedval)
sum(downtown$assessedval > 10000000)

## [1] 16

# Assign test (logical) output to object 'result' by replacing 'FALSE'

result <- downtown$assessedval > 10000000

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Question 5: Total Tax Delinquent Parcels

Question: How many properties are delinquent on taxes? Map your results.

Answer: 57 properties are delinquent on taxes.

# Use function: 'sum()'
# Use variable: 'amtdelinqt'
#summary(downtown$amtdelinqt)
sum(downtown$amtdelinqt > 0)

## [1] 57

# Assign test (logical) output to object 'result' by replacing 'FALSE'

result <- downtown$amtdelinqt > 0

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Question 6: Tax Delinquent Parcels

Question: What proportion of parcels have delinquent tax payments owed?

Answer: 14.65% of parcels have delinquent tax payments owed.

# Use function: 'mean()'
# Use variable: 'amtdelinqt'
group.delinqt <- downtown$amtdelinqt > 0
mean(group.delinqt)*100

## [1] 14.65296

Question 7: Tax Delinquent Commercial Properties

Question I: What proportion of commercial properties are delinquent on taxes?

Question II: What proportion of delinquent tax bills are owed by commercial parcels?

Answer I: 11.96% of commercial properties are delinquent on taxes.

Answer II: 43.86% of delinquent tax dollars are owed by commercial parcels.

# Use function: 'mean()'
# Use variable: 'amtdelinqt'
# Use variable: 'landuse'
group.comdeliqnt <- (downtown$landuse == "Commercial" & downtown$amtdelinqt > 0)
#mean(group.comdeliqnt)
#sum(group.comdeliqnt == TRUE)

# The first answer is tax-delinquent commercial properties over all commercial properties
total1<-sum(downtown$landuse=="Commercial")
prop1<-sum(downtown$landuse=="Commercial" & downtown$amtdelinqt >0)
prop1/total1

## [1] 0.1196172

# The second answer is the tax dollars owed by commercial properties (a subset) over all tax dollars owed
total2<-sum(downtown$amtdelinqt > 0)
prop2<-prop1
prop2/total2

## [1] 0.4385965

Question 8: Tax Delinquent Parcels by Land Use

Question: How many of each land use type are delinquent on taxes? Print a table of your results.

# Use function: 'table()'
# Use variable: 'amtdelinqt'
# Use variable: 'landuse'
group.delinqt <- (downtown$amtdelinqt > 0)

table(downtown$landuse, group.delinqt)

##                     group.delinqt
##                      FALSE TRUE
##   Apartment              6    0
##   Commercial           184   25
##   Community Services    15    2
##   Industrial             2    2
##   Parking               62   16
##   Parks                  8    0
##   Recreation             5    0
##   Religious              6    0
##   Schools                4    0
##   Single Family          1    0
##   Utilities              6    0
##   Vacant Land           33   12

————

DELETE THIS LINE & ALL LINES BELOW IF NOT ATTEPTING CHALLENGES

————

Challenge I: Parcels Values Over $1M

Instructions: Map all of the land valued at over $1 million per acre.

# Replace 'FALSE' with your conditional statement
#summary(downtown)

group.value <- downtown$assessedland/downtown$acres
#sum(group.value > 1000000)
result <- group.value > 1000000

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Challenge II: Parcels Valued Less Than $500K

Instructions: Map all of the land valued at less than $500,000 per acre.

#summary(downtown$assessedland)
# Replace 'FALSE' with your conditional statement
#downtown$valperacre <- downtown$assessedland/downtown$acres  
group.value <- downtown$assessedland/downtown$acres

result <- group.value < 500000

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Challenge III: Total Commercial Value

Instructions: What is the total value of all of the commercial parcels in Downtown?

Answer: The total value of all commercial parcels is $317,067,000.

index <- downtown$landuse == "Commercial" 
z <- downtown[index, ]   
sum(z$assessedval)

## [1] 317067000

# Your code here
#sum(downtown$assessedval)
tapply(downtown$assessedval, downtown$landuse, FUN=sum)

##          Apartment         Commercial Community Services         Industrial 
##           38753600          317067000          158780600           15477700 
##            Parking              Parks         Recreation          Religious 
##           72228540            2315900           56290600            5655600 
##            Schools      Single Family          Utilities        Vacant Land 
##           14201500             110000           25803200            6330500

result <- downtown$landuse =="Commercial" 

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)

Challenge IV: Total Non-Commercial Value

Question: What is the total value of all of the non-commercial parcels in Downtown?

Answer: The total value of all non-commercial parcels is $395,947,740.

# Your code here
index <- downtown$landuse != "Commercial" 
z <- downtown[index, ]   
sum(z$assessedval)

## [1] 395947740

result <- downtown$landuse != "Commercial" 

group.colors <- ifelse(test = result, 
                       yes = "firebrick", 
                       no = "gray80")

plot(downtown,
     border = NA, 
     col = group.colors)