Load Libraries
Reading in, exploring, and cleaning dataset
Read in NYC Real Estate Data
nyc_rolling_sales <- read_csv("/Users/nelsonfarrell/Downloads/nyc-rolling-sales.csv")
Get Column Names
[1] "...1" "BOROUGH"
[3] "NEIGHBORHOOD" "BUILDING CLASS CATEGORY"
[5] "TAX CLASS AT PRESENT" "BLOCK"
[7] "LOT" "EASE-MENT"
[9] "BUILDING CLASS AT PRESENT" "ADDRESS"
[11] "APARTMENT NUMBER" "ZIP CODE"
[13] "RESIDENTIAL UNITS" "COMMERCIAL UNITS"
[15] "TOTAL UNITS" "LAND SQUARE FEET"
[17] "GROSS SQUARE FEET" "YEAR BUILT"
[19] "TAX CLASS AT TIME OF SALE" "BUILDING CLASS AT TIME OF SALE"
[21] "SALE PRICE" "SALE DATE"
View first 6 rows of the dataset
# A tibble: 6 × 22
...1 BOROUGH NEIGHBORHOOD `BUILDING CLASS…` `TAX CLASS AT …` BLOCK
<dbl> <dbl> <chr> <chr> <chr> <dbl>
1 4 1 ALPHABET CITY 07 RENTALS - WAL… 2A 392
2 5 1 ALPHABET CITY 07 RENTALS - WAL… 2 399
3 6 1 ALPHABET CITY 07 RENTALS - WAL… 2 399
4 7 1 ALPHABET CITY 07 RENTALS - WAL… 2B 402
5 8 1 ALPHABET CITY 07 RENTALS - WAL… 2A 404
6 9 1 ALPHABET CITY 07 RENTALS - WAL… 2 405
# … with 16 more variables: LOT <dbl>, `EASE-MENT` <lgl>,
# `BUILDING CLASS AT PRESENT` <chr>, ADDRESS <chr>,
# `APARTMENT NUMBER` <chr>, `ZIP CODE` <dbl>,
# `RESIDENTIAL UNITS` <dbl>, `COMMERCIAL UNITS` <dbl>,
# `TOTAL UNITS` <dbl>, `LAND SQUARE FEET` <chr>,
# `GROSS SQUARE FEET` <chr>, `YEAR BUILT` <dbl>,
# `TAX CLASS AT TIME OF SALE` <dbl>, …
Check the variables in the dataset
spec_tbl_df [84,548 × 22] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
$ ...1 : num [1:84548] 4 5 6 7 8 9 10 11 12 13 ...
$ BOROUGH : num [1:84548] 1 1 1 1 1 1 1 1 1 1 ...
$ NEIGHBORHOOD : chr [1:84548] "ALPHABET CITY" "ALPHABET CITY" "ALPHABET CITY" "ALPHABET CITY" ...
$ BUILDING CLASS CATEGORY : chr [1:84548] "07 RENTALS - WALKUP APARTMENTS" "07 RENTALS - WALKUP APARTMENTS" "07 RENTALS - WALKUP APARTMENTS" "07 RENTALS - WALKUP APARTMENTS" ...
$ TAX CLASS AT PRESENT : chr [1:84548] "2A" "2" "2" "2B" ...
$ BLOCK : num [1:84548] 392 399 399 402 404 405 406 407 379 387 ...
$ LOT : num [1:84548] 6 26 39 21 55 16 32 18 34 153 ...
$ EASE-MENT : logi [1:84548] NA NA NA NA NA NA ...
$ BUILDING CLASS AT PRESENT : chr [1:84548] "C2" "C7" "C7" "C4" ...
$ ADDRESS : chr [1:84548] "153 AVENUE B" "234 EAST 4TH STREET" "197 EAST 3RD STREET" "154 EAST 7TH STREET" ...
$ APARTMENT NUMBER : chr [1:84548] NA NA NA NA ...
$ ZIP CODE : num [1:84548] 10009 10009 10009 10009 10009 ...
$ RESIDENTIAL UNITS : num [1:84548] 5 28 16 10 6 20 8 44 15 24 ...
$ COMMERCIAL UNITS : num [1:84548] 0 3 1 0 0 0 0 2 0 0 ...
$ TOTAL UNITS : num [1:84548] 5 31 17 10 6 20 8 46 15 24 ...
$ LAND SQUARE FEET : chr [1:84548] "1633" "4616" "2212" "2272" ...
$ GROSS SQUARE FEET : chr [1:84548] "6440" "18690" "7803" "6794" ...
$ YEAR BUILT : num [1:84548] 1900 1900 1900 1913 1900 ...
$ TAX CLASS AT TIME OF SALE : num [1:84548] 2 2 2 2 2 2 2 2 2 2 ...
$ BUILDING CLASS AT TIME OF SALE: chr [1:84548] "C2" "C7" "C7" "C4" ...
$ SALE PRICE : chr [1:84548] "6625000" "-" "-" "3936272" ...
$ SALE DATE : POSIXct[1:84548], format: "2017-07-19" ...
- attr(*, "spec")=
.. cols(
.. ...1 = col_double(),
.. BOROUGH = col_double(),
.. NEIGHBORHOOD = col_character(),
.. `BUILDING CLASS CATEGORY` = col_character(),
.. `TAX CLASS AT PRESENT` = col_character(),
.. BLOCK = col_double(),
.. LOT = col_double(),
.. `EASE-MENT` = col_logical(),
.. `BUILDING CLASS AT PRESENT` = col_character(),
.. ADDRESS = col_character(),
.. `APARTMENT NUMBER` = col_character(),
.. `ZIP CODE` = col_double(),
.. `RESIDENTIAL UNITS` = col_double(),
.. `COMMERCIAL UNITS` = col_double(),
.. `TOTAL UNITS` = col_double(),
.. `LAND SQUARE FEET` = col_character(),
.. `GROSS SQUARE FEET` = col_character(),
.. `YEAR BUILT` = col_double(),
.. `TAX CLASS AT TIME OF SALE` = col_double(),
.. `BUILDING CLASS AT TIME OF SALE` = col_character(),
.. `SALE PRICE` = col_character(),
.. `SALE DATE` = col_datetime(format = "")
.. )
- attr(*, "problems")=<externalptr>
Change “SALE PRICE” to Numeric Variable
nyc_rolling_sales$`SALE PRICE` <- as.numeric(nyc_rolling_sales$`SALE PRICE`)
Change borough code numbers into borough names “Manhatten”, “Queens”, etc. In the dataset the different boroughs of New York City were represented with numbers 1-5, 1 Manhatten, 2 the Bronx, etc…
nyc_rolling_sales <- mutate(nyc_rolling_sales, BOROUGH = recode(BOROUGH, `1` = "Manhatten",
`2` = "Bronx", `3` = "Brooklyn", `4` = "Queens", `5` = "Staten Island"))
Recode “BUILDING CLASS” to be more descriptive The building classes in the data were coded. Here I will recode R1-R4, essentially all the single unit residential properties in the dataset
nyc_rolling_sales <- mutate(nyc_rolling_sales, `BUILDING CLASS AT PRESENT` = recode(`BUILDING CLASS AT PRESENT`,
`R4` = "Condo Elevator Building", `R3` = "Condo 1 to 3 Stories ",
`R2` = "Condo Walkup", `R1` = "Condo 2 to 10 Unit Building"))
Select columns “BOROUGH”, “NEIGHBORHOOD”, “BUILDING CLASS AT PRESENT”, “RESIDENTIAL UNITS”, “GROSS SQUARE FEET”, “SALE PRICE”, “SALE DATE” Not all of the variables will be useful for my planned analysis, this removes those that I do not plan use
nyc_rolling_sales <-select(nyc_rolling_sales, "BOROUGH", "NEIGHBORHOOD", "BUILDING CLASS AT PRESENT",
"RESIDENTIAL UNITS", "GROSS SQUARE FEET", "SALE PRICE", "SALE DATE")
Rename columns “BUILDING CLASS AT PRESENT”, and “RESIDENTIAL UNITS” This changes the names of some of the columns for clarity, “Residential Units” will become “Units” because I will only be looking at Residential Units. “BUILDING CLASS AT PRESENT” will become simply "BUILDING CLASS
nyc_rolling_sales <- rename(nyc_rolling_sales, "BUILDING CLASS" = "BUILDING CLASS AT PRESENT",
"SQ FOOTAGE" = "GROSS SQUARE FEET", "UNITS" = "RESIDENTIAL UNITS")
Remove all observations that do not contain a residential unit, and all sale prices under $100,000 and over $5,000,000 There were a lot very low prices that according to data description could indicate family member to family member transfer, i.e., inheritance. The data is also heavily skewed to the right, so I will look at only specific range of prices (i.e., $100,000-$5,000,000)
nyc_rolling_sales <- filter(nyc_rolling_sales, ! `UNITS` == "0" & ! `SALE PRICE` == "-" &
`SALE PRICE` > 100000 & `SALE PRICE` < 5000000)
arrange(nyc_rolling_sales, `SALE PRICE`)
# A tibble: 37,897 × 7
BOROUGH NEIGHBORHOOD `BUILDING CLASS` UNITS `SQ FOOTAGE`
<chr> <chr> <chr> <dbl> <chr>
1 Staten Island MIDLAND BEACH "A6" 1 588
2 Brooklyn PARK SLOPE SOUTH "C0" 3 2781
3 Queens FLUSHING-SOUTH "Condo 1 to 3 S… 1 -
4 Queens JAMAICA "S2" 2 3225
5 Bronx MELROSE/CONCOURSE "C0" 3 3972
6 Brooklyn BOERUM HILL "Condo Elevator… 1 0
7 Bronx PARKCHESTER "Condo Elevator… 1 0
8 Bronx PARKCHESTER "Condo Elevator… 1 0
9 Queens REGO PARK "B3" 2 2128
10 Bronx PARKCHESTER "Condo Elevator… 1 0
# … with 37,887 more rows, and 2 more variables: `SALE PRICE` <dbl>,
# `SALE DATE` <dttm>
# A tibble: 6 × 7
BOROUGH NEIGHBORHOOD `BUILDING CLASS` UNITS `SQ FOOTAGE`
<chr> <chr> <chr> <dbl> <chr>
1 Manhatten ALPHABET CITY C4 10 6794
2 Manhatten ALPHABET CITY C4 8 4226
3 Manhatten ALPHABET CITY Condo Walkup 1 -
4 Manhatten ALPHABET CITY Condo Elevator Building 1 -
5 Manhatten ALPHABET CITY Condo Elevator Building 1 -
6 Manhatten ALPHABET CITY Condo Elevator Building 1 -
# … with 2 more variables: `SALE PRICE` <dbl>, `SALE DATE` <dttm>
Identify and view the different types of building classes Each of these identifiers is defined in the description of the dataset, R4s (“Condo; Residential Unit in Elevator Building”) are the most numerous. It will be most revealing to compare like dwellings so I have renamed R1-R4, and I will filter out the rest
tally(~ `BUILDING CLASS`, data = nyc_rolling_sales)
BUILDING CLASS
A0 A1
307 4666
A2 A3
1895 238
A4 A5
91 4023
A6 A7
62 2
A9 B1
949 2718
B2 B3
3124 2437
B9 C0
918 2245
C1 C2
267 369
C3 C4
538 46
C5 C6
63 7
C7 C9
66 5
Condo 1 to 3 Stories Condo 2 to 10 Unit Building
1076 934
Condo Elevator Building Condo Walkup
8931 646
D1 D3
11 3
D4 D6
24 6
D7 D9
2 2
E9 F9
2 2
G0 G4
7 1
G9 HR
1 2
I7 K1
2 2
K2 K4
2 99
K9 M1
3 1
M9 N2
5 1
N9 O2
2 1
O5 O7
2 1
O8 P5
18 2
R6 RR
129 9
S0 S1
7 202
S2 S3
403 79
S4 S5
61 50
S9 V0
96 28
V1 W2
3 1
Z0 Z9
1 1
Filter “Building Class” for “Condo Elevator Building”, previously “R4” Here I have filtered for just “Condo Elevator Building”
nyc_rolling_sales<- filter(nyc_rolling_sales, `BUILDING CLASS` == "Condo Elevator Building")
head(nyc_rolling_sales)
# A tibble: 6 × 7
BOROUGH NEIGHBORHOOD `BUILDING CLASS` UNITS `SQ FOOTAGE`
<chr> <chr> <chr> <dbl> <chr>
1 Manhatten ALPHABET CITY Condo Elevator Building 1 -
2 Manhatten ALPHABET CITY Condo Elevator Building 1 -
3 Manhatten ALPHABET CITY Condo Elevator Building 1 -
4 Manhatten ALPHABET CITY Condo Elevator Building 1 -
5 Manhatten ALPHABET CITY Condo Elevator Building 1 -
6 Manhatten ALPHABET CITY Condo Elevator Building 1 -
# … with 2 more variables: `SALE PRICE` <dbl>, `SALE DATE` <dttm>
The Variables in the dataset after cleaning the data
- BOROUGH: Character, this identifies which borough of New York City the property is in.
- NEIGHBORHOOD: Character, this identifies the specific neighborhood within each borough that the property is in.
- BUILDING CLASS: Character, this identifies what type of building the property is.
- UNITS: Numeric, this identifies the number of units in the property
- SQ FOOTAGE: Character, still currently character although I will change it numeric if it turns out to be useful. Displays the total area of the property, inside and out, in square feet
- SALE PRICE: Numeric, this is the sale price of the property
- SALE DATE: Numeric, this is the date the property was sold.
Exploratory analyiss
Create Object “Brooklyn_1” for property sales of “CONDO ELEVATOR BUILDINGs” in Brooklyn
Brooklyn_1 <- filter(nyc_rolling_sales, `BOROUGH` == "Brooklyn")
Create Object “Manhatten_1” for property sales of “CONDO ELEVATOR BUILDINGs” in Manhatten
Manhatten_1 <- filter(nyc_rolling_sales, `BOROUGH` == "Manhatten")
Display Boxplot of “SALE PRICE” of “CONDO ELEVATOR BUILDINGS” in Brooklyn & Manhatten Here we can still clearly see that even after looking at only a range of the “SALE PRICE” the data is still heavily skewed to the right with numerous outliers. As a result, I will use median as a measure of center.
options(scipen = 999) ##Remove Scientific Notation
boxplot(Brooklyn_1$`SALE PRICE`, Manhatten_1$`SALE PRICE`, horizontal = TRUE, names=c("Brooklyn", "Manhatten"),
main = "Sales of Single Unit Homes in Brooklyn & Manhatten", ylab = "Boroughs", xlab = "Sale Price")
Display summary statistics of “Brooklyn_1” and “SALE PRICE” and “Manhatten_1” and “SALE PRICE”
min Q1 median Q3 max mean sd n missing
101825 682227 936270 1482500 4938512 1206647 806376.9 2159 0
min Q1 median Q3 max mean sd n missing
150000 900000 1396500 2310000 4996841 1734470 1074408 4783 0
Find the median “SALE PRICE” of “CONDO ELEVATOR BUILDINGs” in each of the five New York City “BOROUGHS”
# A tibble: 5 × 2
BOROUGH `SALE PRICE`
<chr> <dbl>
1 Bronx 143500
2 Brooklyn 936270
3 Manhatten 1396500
4 Queens 643000
5 Staten Island 240318
Find the median “SALE PRICE” of “CONDO ELEVATOR BUILDINGs” in each of the different Manhatten “NEIGHBORHOODs”
# A tibble: 38 × 2
NEIGHBORHOOD `SALE PRICE`
<chr> <dbl>
1 ALPHABET CITY 1310000
2 CHELSEA 1820000
3 CHINATOWN 1440000
4 CIVIC CENTER 3250000
5 CLINTON 1236250
6 EAST VILLAGE 1285000
7 FASHION 1735000
8 FINANCIAL 1450000
9 FLATIRON 2300000
10 GRAMERCY 1834270.
# … with 28 more rows
Potential Research Questions
- Which BOROUGH has the highest median property value?
- Which NEIGHBORHOODs within each BOROUGH have the highest median property value?
- Which BOROUGH’s median property value is increasing (decreasing) the fastest?
- In which BOROUGH would it make the most financial sense to purchase an investment property?
- Which NEIGHBORHOOD’s property value is increasing (decreasing) the fastest?
- In which NEIGHBORHOOD would it make the most financial sense to purchase an investment property?
- Are there correlations between property values in BOROUGHs and/or NEIGHBORHOODs?
- Determine and examine price per square foot as a metric. (Unclear if this is possible as this data is missing from many of the observations)
Reuse
Text and figures are licensed under Creative Commons Attribution CC BY-NC 4.0. The figures that have been reused from other sources don't fall under this license and can be recognized by a note in their caption: "Figure from ...".
Citation
For attribution, please cite this work as
Farrell (2022, Feb. 16). Data Analytics and Computational Social Science: Homework3. Retrieved from https://github.com/DACSS/dacss_course_website/posts/httpsrpubscomjnfarrell211866926/
BibTeX citation
@misc{farrell2022homework3,
author = {Farrell, Joseph},
title = {Data Analytics and Computational Social Science: Homework3},
url = {https://github.com/DACSS/dacss_course_website/posts/httpsrpubscomjnfarrell211866926/},
year = {2022}
}
