library(googleway)
library(tidytransit)
library(measurements)
library(lubridate)
library(sf)
library(tidyverse)Using Google Maps via R to explore Greater Boston neighborhood housing
Rendered 07/09/2024
Background
We’d like to explore housing in the greater Boston area, but given the high prices and low inventory, the number of neighborhoods each with their own characteristics, and the tricky fact that our on-site work locations are in different areas, we’d like a way to streamline our search. The goal here is to locate areas from which we can still reliably commute via public transit but also be walking distance to neighborhood points-of-interest important to us. From there, we can find housing listings that match various other search criteria.
Some of this was inspired by a post here demonstrating that plotting commute times via Google Maps API calls in R was possible. I wanted to add my own transit-connectivity index, based on the fact that MBTA is somewhat unreliable (thus having potential redundancy in our commute methods would be beneficial), and then compute an overall walkability score based on this and other highly rated points-of-interest within walking distance. The post linked above tiled hexagons across a map of Melbourne; instead I decided to use 2020 census blocks, and their centroids, as the unit on which most of this would be computed.
Much of this is possible thanks to two R packages: googleway and sf. Things like calculating distances, computing centroids, intersections, etc of spatial geometries are all possible in sf and then interactions with the Google Maps API (e.g. getting maps navigation, or searching for points of interest) are carried out via googleway. ggmap is also worth a look as it makes maps functionality work within ggplot2.
Lastly, I leveraged the python package HomeHarvest to retrieve all recent MLS listings in the area and overlay these as points on our map
Note and disclaimer: Google API calls can add up quickly. When you register for an API key, review all pricing details and decide for yourself whether and how to proceed so as to not generate a massive bill. All maps images here are static, but when you plot via googleway::google_map() they will be interactive.
Initial setup
Load dependencies
Set private Google Maps API
I’ve set mine as an environment variable so I retrieve that value here
#api <- "your_private_api_key_here"
api <- Sys.getenv("GOOGLE_API")
googleway::set_key(key = api)Read Boston MBTA data via GTFS
boston_mbta <- read_gtfs("https://cdn.mbta.com/MBTA_GTFS.zip")Warning in gtfs_to_tidygtfs(g, files = files): Duplicated ids found in: fare_leg_rules
The returned object is not a tidygtfs object, you can use as_tidygtfs() after fixing the issue.Set reasonable walking distances for points of interest (0.75mi) and transit (0.5mi) to meters
max_walk_meters <- floor(measurements::conv_unit(x = 0.75,from = "mi",to="m"))
max_walk_meters_transport <- floor(measurements::conv_unit(x = 0.5,from = "mi",to="m"))Prepare and tidy up MBTA transit data
Get all MBTA RapidTransit stops in the network, separated by T, CR, and bus
t_stops <- boston_mbta$stops %>%
dplyr::filter(zone_id == "RapidTransit") %>%
dplyr::select(stop_id,stop_name,stop_lat,stop_lon) %>%
dplyr::distinct(stop_id, .keep_all = TRUE)
t_stops stop_id stop_name stop_lat stop_lon
1: 70061 Alewife 42.39616 -71.13997
2: Alewife-01 Alewife 42.39616 -71.13997
3: Alewife-02 Alewife 42.39616 -71.13997
4: 70126 Allston Street 42.34865 -71.13788
5: 70127 Allston Street 42.34855 -71.13736
---
273: 70060 Wonderland 42.41336 -70.99168
274: 70162 Woodland 42.33270 -71.24305
275: 70163 Woodland 42.33309 -71.24366
276: 70124 Warren Street 42.34828 -71.14044
277: 70125 Warren Street 42.34882 -71.14005cr_stops <- boston_mbta$stops %>%
dplyr::filter(zone_id == "CR-zone-1A") %>%
dplyr::select(stop_id,stop_name,stop_lat,stop_lon) %>%
dplyr::distinct(stop_id, .keep_all = TRUE)
cr_stops stop_id stop_name stop_lat stop_lon
1: place-DB-2205 Fairmount 42.25364 -71.11927
2: DB-2205 Fairmount 42.25364 -71.11927
3: DB-2205-01 Fairmount 42.25364 -71.11927
4: DB-2205-02 Fairmount 42.25364 -71.11927
5: place-DB-2222 Blue Hill Avenue 42.27147 -71.09578
---
152: NEC-2287-09 South Station 42.35114 -71.05508
153: NEC-2287-10 South Station 42.35103 -71.05496
154: NEC-2287-11 South Station 42.35103 -71.05496
155: NEC-2287-12 South Station 42.35074 -71.05493
156: NEC-2287-13 South Station 42.35074 -71.05493bus_stops <- boston_mbta$stops %>%
dplyr::filter(zone_id == "LocalBus" | zone_id == "ExpressBus-Downtown") %>%
dplyr::select(stop_id,stop_name,stop_lat,stop_lon) %>%
dplyr::distinct(stop_id, .keep_all = TRUE)
bus_stops stop_id stop_name stop_lat stop_lon
1: 1 Washington St opp Ruggles St 42.33096 -71.08275
2: 10 Theo Glynn Way @ Newmarket Sq 42.33055 -71.06879
3: 10000 Tremont St opp Temple Pl 42.35569 -71.06291
4: 10003 Albany St opp Randall St 42.33159 -71.07624
5: 10005 Albany St opp E Concord St 42.33502 -71.07128
---
6771: 29013 Sullivan Square 42.38403 -71.07609
6772: 29014 Sullivan Square 42.38392 -71.07609
6773: 5271 Wellington Station Busway 42.40258 -71.07609
6774: 5666 Wood Island 42.38037 -71.02329
6775: 15795 Wonderland 42.41338 -70.99205Create lookup function to pull together stop names and lines
Which transit stops are on which lines is not as easily linked as I expected, though it is all present in the GTFS feed. I perform a series of joins using the logic described here, though my approach was a bit different
stop_route_lookup <- full_join(boston_mbta$routes,boston_mbta$trips) %>%
full_join(boston_mbta$stop_times) %>%
full_join(boston_mbta$stops) %>%
dplyr::select(trip_id, route_id, shape_id, stop_id) %>%
distinct()Joining with `by = join_by(route_id)`
Joining with `by = join_by(trip_id)`
Joining with `by = join_by(stop_id)`Annotate T stops with Lines, then create walking distance buffer around each stop
t_stops_withLines <- full_join(t_stops,stop_route_lookup,by="stop_id") %>%
dplyr::select(-stop_id,-trip_id,-shape_id) %>%
as_tibble() %>%
drop_na(route_id) %>%
drop_na(stop_lat) %>%
dplyr::filter(route_id != "Shuttle") %>%
distinct()
t_stops_withLines_buffered <- st_as_sf(t_stops_withLines, coords = c("stop_lon","stop_lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters_transport)
t_stops_withLines_bufferedSimple feature collection with 293 features and 2 fields
Geometry type: POLYGON
Dimension: XY
Bounding box: xmin: -71.26209 ymin: 42.20009 xmax: -70.9818 ymax: 42.44505
Geodetic CRS: WGS 84
# A tibble: 293 × 3
stop_name route_id geometry
* <chr> <chr> <POLYGON [°]>
1 Alewife Red ((-71.14487 42.40244, -71.14487 42.40242, -71.14504 …
2 Allston Street Green-B ((-71.13371 42.35524, -71.1339 42.35527, -71.13399 4…
3 Allston Street Green-B ((-71.13371 42.35531, -71.1338 42.35533, -71.1339 42…
4 Amory Street Green-B ((-71.10902 42.35698, -71.10902 42.35705, -71.10911 …
5 Amory Street Green-B ((-71.1092 42.3569, -71.1092 42.35701, -71.10939 42.…
6 Andrew Red ((-71.04968 42.32591, -71.04959 42.32589, -71.04959 …
7 Airport Blue ((-71.03789 42.36961, -71.03789 42.36948, -71.03771 …
8 Aquarium Blue ((-71.06092 42.36243, -71.06092 42.36228, -71.06092 …
9 Arlington Green-B ((-71.06157 42.34955, -71.06157 42.34962, -71.06148 …
10 Arlington Green-C ((-71.06157 42.34955, -71.06157 42.34962, -71.06148 …
# ℹ 283 more rowsAnnotate CR stops with Lines, then create walking distance buffer around each stop
cr_stops_withLines <- full_join(cr_stops,stop_route_lookup,by="stop_id") %>%
dplyr::select(-stop_id,-trip_id,-shape_id) %>%
as_tibble() %>%
drop_na(route_id) %>%
drop_na(stop_lat) %>%
dplyr::filter(route_id != "Shuttle") %>%
distinct()
cr_stops_withLines_buffered <- st_as_sf(cr_stops_withLines, coords = c("stop_lon","stop_lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters_transport)
cr_stops_withLines_bufferedSimple feature collection with 94 features and 2 fields
Geometry type: POLYGON
Dimension: XY
Bounding box: xmin: -71.14968 ymin: 42.24633 xmax: -70.61722 ymax: 42.80517
Geodetic CRS: WGS 84
# A tibble: 94 × 3
stop_name route_id geometry
* <chr> <chr> <POLYGON [°]>
1 Fairmount CR-Fairmount ((-71.12898 42.25275, -71.12898 42.25268, -71…
2 Fairmount CR-Franklin ((-71.12898 42.25275, -71.12898 42.25268, -71…
3 Fairmount CR-Providence ((-71.12898 42.25275, -71.12898 42.25268, -71…
4 Blue Hill Avenue CR-Fairmount ((-71.10458 42.27465, -71.10458 42.27463, -71…
5 Blue Hill Avenue CR-Franklin ((-71.10458 42.27465, -71.10458 42.27463, -71…
6 Blue Hill Avenue CR-Providence ((-71.10458 42.27465, -71.10458 42.27463, -71…
7 Morton Street CR-Fairmount ((-71.08506 42.27376, -71.08497 42.27374, -71…
8 Morton Street CR-Franklin ((-71.08506 42.27376, -71.08497 42.27374, -71…
9 Morton Street CR-Providence ((-71.08506 42.27376, -71.08497 42.27374, -71…
10 Talbot Avenue CR-Fairmount ((-71.08692 42.28906, -71.08692 42.28887, -71…
# ℹ 84 more rowsAnnotate bus stops with Lines, then create walking distance buffer around each stop
bus_stops_withLines <- full_join(bus_stops,stop_route_lookup,by="stop_id") %>%
dplyr::select(-stop_id,-trip_id,-shape_id) %>%
as_tibble() %>%
drop_na(route_id) %>%
drop_na(stop_lat) %>%
distinct()
bus_stops_withLines_buffered <- st_as_sf(bus_stops_withLines, coords = c("stop_lon","stop_lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters_transport)
bus_stops_withLines_bufferedSimple feature collection with 11255 features and 2 fields
Geometry type: POLYGON
Dimension: XY
Bounding box: xmin: -71.30106 ymin: 42.09942 xmax: -70.83522 ymax: 42.59434
Geodetic CRS: WGS 84
# A tibble: 11,255 × 3
stop_name route_id geometry
* <chr> <chr> <POLYGON [°]>
1 Washington St opp Ruggles St 1 ((-71.07345 42.32856, -71.07345 42.32…
2 Washington St opp Ruggles St 8 ((-71.07345 42.32856, -71.07345 42.32…
3 Washington St opp Ruggles St 15 ((-71.07345 42.32856, -71.07345 42.32…
4 Washington St opp Ruggles St 19 ((-71.07345 42.32856, -71.07345 42.32…
5 Washington St opp Ruggles St 47 ((-71.07345 42.32856, -71.07345 42.32…
6 Washington St opp Ruggles St 171 ((-71.07345 42.32856, -71.07345 42.32…
7 Theo Glynn Way @ Newmarket Sq 8 ((-71.06798 42.3378, -71.06816 42.337…
8 Theo Glynn Way @ Newmarket Sq 10 ((-71.06798 42.3378, -71.06816 42.337…
9 Theo Glynn Way @ Newmarket Sq 171 ((-71.06798 42.3378, -71.06816 42.337…
10 Tremont St opp Temple Pl 15 ((-71.05497 42.35147, -71.05479 42.35…
# ℹ 11,245 more rowsPrepare census block information. We’ll use this as a means of creating tiles around the greater Boston area
Import Massachusetts 2020 census blocks
mass_censusblocks <- st_read(dsn = "~/Downloads/CENSUS2020_BLK_BG_TRCT/CENSUS2020BLOCKS_POLY.shp")Reading layer `CENSUS2020BLOCKS_POLY' from data source
`/Users/jsimon/Downloads/CENSUS2020_BLK_BG_TRCT/CENSUS2020BLOCKS_POLY.shp'
using driver `ESRI Shapefile'
Simple feature collection with 106164 features and 22 fields
Geometry type: MULTIPOLYGON
Dimension: XY
Bounding box: xmin: 33863.75 ymin: 777634.4 xmax: 330838.8 ymax: 959743
Projected CRS: NAD83 / Massachusetts MainlandCreate a list of Greater Boston area towns
greater_boston_towns <- c(
"Arlington",
"Belmont",
"Boston",
"Braintree",
"Brookline",
"Cambridge",
"Chelsea",
"Everett",
"Malden",
"Medford",
"Melrose",
"Milton",
"Newton",
"Quincy",
"Revere",
"Somerville",
"Waltham",
"Watertown",
"Winthrop"
)Filter census blocks for those in Greater Boston area towns
greaterBoston_censusblocks <- mass_censusblocks[mass_censusblocks$TOWN %in% toupper(greater_boston_towns),]
# A handful of these have invalid boundaries and will cause errors later, so let's repair them
greaterBoston_censusblocks <- st_make_valid(greaterBoston_censusblocks)
# Several blocks fall over water or other unpopulated areas, so let's remove those by asserting that population as of the 2020 census has to be non-zero
greaterBoston_censusblocks <- greaterBoston_censusblocks[greaterBoston_censusblocks$POP20 > 0,]
greaterBoston_censusblocksSimple feature collection with 14449 features and 22 fields
Geometry type: POLYGON
Dimension: XY
Bounding box: xmin: 217614.4 ymin: 879882.7 xmax: 245421.3 ymax: 914028.5
Projected CRS: NAD83 / Massachusetts Mainland
First 10 features:
STATEFP20 COUNTYFP20 TRACTCE20 BLOCKCE20 GEOID20 NAME20
51198 25 025 180101 2008 250251801012008 Block 2008
51199 25 025 130300 3007 250251303003007 Block 3007
51200 25 025 050700 1004 250250507001004 Block 1004
51201 25 025 160300 2002 250251603002002 Block 2002
51203 25 021 419400 2012 250214194002012 Block 2012
51204 25 021 400700 3002 250214007003002 Block 3002
51205 25 021 418101 1017 250214181011017 Block 1017
51206 25 021 417704 2003 250214177042003 Block 2003
51208 25 021 419602 2021 250214196022021 Block 2021
51210 25 021 416101 3012 250214161013012 Block 3012
MTFCC20 ALAND20 AWATER20 INTPTLAT20 INTPTLON20 HOUSING20 POP20
51198 G5040 12243 0 +42.3885921 -070.9729859 23 57
51199 G5040 10029 0 +42.2786967 -071.1515725 12 32
51200 G5040 14078 0 +42.3735839 -071.0339670 121 354
51201 G5040 31455 0 +42.3897089 -071.0452321 315 370
51203 G5040 7006 0 +42.2291968 -070.9697674 8 30
51204 G5040 15162 0 +42.3339984 -071.1275203 27 51
51205 G5040 21732 0 +42.2539706 -071.0083784 38 113
51206 G5040 29460 0 +42.2549821 -071.0044867 111 252
51208 G5040 20846 0 +42.1739325 -071.0084655 13 40
51210 G5040 17006 0 +42.2334696 -071.0691404 15 54
AREA_SQFT AREA_ACRES TOWN TOWN_ID BLKGRP20 TRACT20 COUSUBFP
51198 131772.95 3.03 WINTHROP 346 250251801012 25025180101 81005
51199 107941.62 2.48 BOSTON 35 250251303003 25025130300 07000
51200 151520.76 3.48 BOSTON 35 250250507001 25025050700 07000
51201 338558.53 7.77 CHELSEA 57 250251603002 25025160300 13205
51203 75409.17 1.73 BRAINTREE 40 250214194002 25021419400 07740
51204 163192.32 3.75 BROOKLINE 46 250214007003 25021400700 09175
51205 233902.16 5.37 QUINCY 243 250214181011 25021418101 55745
51206 317077.43 7.28 QUINCY 243 250214177042 25021417704 55745
51208 224364.62 5.15 BRAINTREE 40 250214196022 25021419602 07740
51210 183032.73 4.20 MILTON 189 250214161013 25021416101 41690
SHAPE_AREA SHAPE_LEN geometry
51198 12242.156 543.5588 POLYGON ((243402.5 904425.5...
51199 10028.145 416.8513 POLYGON ((228812 892112.3, ...
51200 14076.795 556.6098 POLYGON ((238477.5 902727.8...
51201 31453.243 701.0596 POLYGON ((237399.1 904332.8...
51203 7005.769 393.1014 POLYGON ((243842.7 886675.6...
51204 15161.123 519.5947 POLYGON ((230792.6 898286.1...
51205 21730.308 730.1781 POLYGON ((240694.7 889295.9...
51206 29457.575 827.4635 POLYGON ((240985.5 889522, ...
51208 20844.239 633.6230 POLYGON ((240690.1 880524.3...
51210 17004.365 591.8149 POLYGON ((235666.6 887046, ...Intersect census blocks with MBTA transit data to find which blocks have walkable T, CR, and bus routes
Calculate which census blocks intersect buffered T, CR, and bus stops
blocks_walkToT <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,t_stops_withLines_buffered)
blocks_walkToTSparse geometry binary predicate list of length 14449, where the
predicate was `intersects'
first 10 elements:
1: (empty)
2: (empty)
3: 7, 207
4: (empty)
5: (empty)
6: 34, 39, 40, 53, 54, 65, 113, 114, 267, 268
7: 232
8: 232
9: (empty)
10: (empty)blocks_walkToCR <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,cr_stops_withLines_buffered)
blocks_walkToCRSparse geometry binary predicate list of length 14449, where the
predicate was `intersects'
first 10 elements:
1: (empty)
2: (empty)
3: (empty)
4: 57, 58
5: (empty)
6: (empty)
7: (empty)
8: (empty)
9: (empty)
10: (empty)blocks_walkToBus <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,bus_stops_withLines_buffered)
blocks_walkToBusSparse geometry binary predicate list of length 14449, where the
predicate was `intersects'
first 10 elements:
1: 223, 224, 225, 226, 227, 268, 269, 270, 271, 290, ...
2: 600, 6255, 6256, 6257, 6258, 6259, 6307, 6308, 6309, 6311, ...
3: 1416, 6962, 7069, 7070, 7071, 7072, 7073, 7074, 7075, 7076, ...
4: 1197, 1198, 1199, 1200, 1201, 1202, 1203, 1204, 1205, 1206, ...
5: 4459, 4460, 4461, 4462, 4463, 4464, 4465, 4467, 4497, 4498, ...
6: 787, 788, 791, 795, 796, 798, 805, 817, 822, 823, ...
7: 849, 893, 1046, 2522, 3529, 3530, 3531, 3532, 3533, 3534, ...
8: 3529, 3530, 3531, 3532, 3533, 3534, 3538, 3539, 3540, 3541, ...
9: 4637, 4638, 4639, 4640, 4641, 4642, 4643, 4658, 4659, 4660, ...
10: 4963, 4964, 4965, 4966, 4972, 4973, 4975, 4976, 5189, 5190, ...Summarize number of unique T lines (by color) per census block
Here we collapse all branches (-B, -C, -D, -E) of green, and call “Mattapan” as “Red” for simplicity
Note this penalizes the Green line somewhat, given the distinct branches west of Kenmore. However given how slow the western branches of the Green line are, this score prioritizes having some truly rapid option and/or other mode of transit
nLines_perBlock <- t_stops_withLines_buffered %>%
as_tibble() %>%
dplyr::select(-geometry) %>%
mutate(RowNumber = row_number()) %>%
mutate(route_id = str_replace_all(route_id,"-.","")) %>%
mutate(route_id = str_replace_all(route_id,"Mattapan","Red")) %>%
left_join(enframe(blocks_walkToT) %>%
unnest(cols=value,keep_empty = T),.,by=c("value" = "RowNumber")) %>%
group_by(name) %>%
summarize(nLines = n_distinct(route_id,na.rm=T))
nLines_perBlock# A tibble: 14,449 × 2
name nLines
<int> <int>
1 1 0
2 2 0
3 3 1
4 4 0
5 5 0
6 6 1
7 7 1
8 8 1
9 9 0
10 10 0
# ℹ 14,439 more rows# Add this annotation to our census blocks as a new column
greaterBoston_censusblocks$numTlines <- nLines_perBlock$nLines Summarize number of unique CR lines per census block
nCRLines_perBlock <- cr_stops_withLines_buffered %>%
as_tibble() %>%
dplyr::select(-geometry) %>%
mutate(RowNumber = row_number()) %>%
left_join(enframe(blocks_walkToCR) %>%
unnest(cols=value,keep_empty = T),.,by=c("value" = "RowNumber")) %>%
group_by(name) %>%
summarize(nLines = n_distinct(route_id,na.rm=T))
nCRLines_perBlock# A tibble: 14,449 × 2
name nLines
<int> <int>
1 1 0
2 2 0
3 3 0
4 4 1
5 5 0
6 6 0
7 7 0
8 8 0
9 9 0
10 10 0
# ℹ 14,439 more rows# Add this annotation to our census blocks as a new column
greaterBoston_censusblocks$numCRlines <- nCRLines_perBlock$nLines Summarize number of unique bus lines per census block
nBusLines_perBlock <- bus_stops_withLines_buffered %>%
as_tibble() %>%
dplyr::select(-geometry) %>%
mutate(RowNumber = row_number()) %>%
left_join(enframe(blocks_walkToBus) %>%
unnest(cols=value,keep_empty = T),.,by=c("value" = "RowNumber")) %>%
group_by(name) %>%
summarize(nLines = n_distinct(route_id,na.rm=T))
nBusLines_perBlock# A tibble: 14,449 × 2
name nLines
<int> <int>
1 1 2
2 2 4
3 3 10
4 4 5
5 5 1
6 6 3
7 7 18
8 8 18
9 9 1
10 10 1
# ℹ 14,439 more rows# Add this annotation to our census blocks as a new column
greaterBoston_censusblocks$numBuslines <- nBusLines_perBlock$nLines Compute aggregated transit connectivity score per census block and plot our first map
Scored as:
0: no T, CR, or bus
1: only one of T/CR/bus (ie relies only on one form of transit)
2: T and either CR or bus present
3: T, CR, bus all present
T_connectivity <- cbind("numTlines" = greaterBoston_censusblocks$numTlines,
"numCRlines" = greaterBoston_censusblocks$numCRlines,
"numBuslines" = greaterBoston_censusblocks$numBuslines
) %>%
as_tibble() %>%
mutate(Connectivity = case_when(
(numTlines == 0) & (numCRlines == 0) & (numBuslines == 0) ~ 0,
(numTlines > 0) & (numCRlines > 0) & (numBuslines > 0) ~ 3,
(numTlines > 0) & ((numCRlines > 0) | (numBuslines > 0)) ~ 2,
(numCRlines > 0) & ((numTlines > 0) | (numBuslines > 0)) ~ 2,
((numTlines == 0) & (numCRlines == 0)) |
((numTlines == 0) & (numBuslines == 0)) |
((numCRlines == 0) & (numBuslines == 0)) ~ 1,
T ~ 999
)
)
T_connectivity# A tibble: 14,449 × 4
numTlines numCRlines numBuslines Connectivity
<int> <int> <int> <dbl>
1 0 0 2 1
2 0 0 4 1
3 1 0 10 2
4 0 1 5 2
5 0 0 1 1
6 1 0 3 2
7 1 0 18 2
8 1 0 18 2
9 0 0 1 1
10 0 0 1 1
# ℹ 14,439 more rows# Check that all scored as expected
sum(T_connectivity$Connectivity==999)[1] 0# Add Connectivity score to census block data matrix
greaterBoston_censusblocks$T_connectivity <- T_connectivity$ConnectivityPlot base map then color census blocks by T connectivity score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(greaterBoston_censusblocks,crs = 4326),
fill_colour = "T_connectivity",
fill_opacity = 0.75,
legend = T) 
Create a database of highly-rated points-of-interest that we factor in for good livability
Here we include cafes, restaurants, parks/playgrounds, and grocery stores. I performed these queries more by neighborhood than town, since Google Maps retrieves results regardless for a fairly large area
Retrieve listings for highly-reviewed cafes in the greater Boston area
Note the “coffee roaster” term is used here to try to avoid Dunkin’, Starbucks, etc, but may miss some other good-quality cafes. Your priorities may vary.
The google_places search can only retrieve 20 results at a time, so we grab a few pages worth for each neighborhood, then merge everything together, set a filter for star rating and a minimum number of reviews, and remove duplicates
cafes_somerville <- google_places(search_string = "Coffee roaster in Somerville, MA",
key = api)
cafes_somerville_next1 <- google_places(search_string = "Coffee roaster in Somerville, MA",
page_token = cafes_somerville$next_page_token,
key = api)
cafes_somerville_next2 <- google_places(search_string = "Coffee roaster in Somerville, MA",
page_token = cafes_somerville_next1$next_page_token,
key = api)
cafes_cambridge <- google_places(search_string = "Coffee roaster in Cambridge, MA",
key = api)
cafes_cambridge_next1 <- google_places(search_string = "Coffee roaster in Cambridge, MA",
page_token = cafes_cambridge$next_page_token,
key = api)
cafes_cambridge_next2 <- google_places(search_string = "Coffee roaster in Cambridge, MA",
page_token = cafes_cambridge_next1$next_page_token,
key = api)
cafes_boston <- google_places(search_string = "Coffee roaster in Boston, MA",
key = api)
cafes_boston_next1 <- google_places(search_string = "Coffee roaster in Boston, MA",
page_token = cafes_boston$next_page_token,
key = api)
cafes_boston_next2 <- google_places(search_string = "Coffee roaster in Boston, MA",
page_token = cafes_boston_next1$next_page_token,
key = api)
cafes_brookline <- google_places(search_string = "Coffee roaster in Brookline, MA",
key = api)
cafes_brookline_next1 <- google_places(search_string = "Coffee roaster in Brookline, MA",
page_token = cafes_brookline$next_page_token,
key = api)
cafes_brookline_next2 <- google_places(search_string = "Coffee roaster in Brookline, MA",
page_token = cafes_brookline_next1$next_page_token,
key = api)
cafes_jp <- google_places(search_string = "Coffee roaster in Jamaica Plain, MA",
key = api)
cafes_jp_next1 <- google_places(search_string = "Coffee roaster in Jamaica Plain, MA",
page_token = cafes_jp$next_page_token,
key = api)
cafes_jp_next2 <- google_places(search_string = "Coffee roaster in Jamaica Plain, MA",
page_token = cafes_jp_next1$next_page_token,
key = api)
cafes_medford <- google_places(search_string = "Coffee roaster in Medford, MA",
key = api)
cafes_medford_next1 <- google_places(search_string = "Coffee roaster in Medford, MA",
page_token = cafes_medford$next_page_token,
key = api)
cafes_medford_next2 <- google_places(search_string = "Coffee roaster in Medford, MA",
page_token = cafes_medford_next1$next_page_token,
key = api)
cafes_belmont <- google_places(search_string = "Coffee roaster in Belmont, MA",
key = api)
cafes_belmont_next1 <- google_places(search_string = "Coffee roaster in Belmont, MA",
page_token = cafes_belmont$next_page_token,
key = api)
cafes_belmont_next2 <- google_places(search_string = "Coffee roaster in Belmont, MA",
page_token = cafes_belmont_next1$next_page_token,
key = api)
cafes_dorchester <- google_places(search_string = "Coffee roaster in Dorchester, MA",
key = api)
cafes_dorchester_next1 <- google_places(search_string = "Coffee roaster in Dorchester, MA",
page_token = cafes_dorchester$next_page_token,
key = api)
cafes_dorchester_next2 <- google_places(search_string = "Coffee roaster in Dorchester, MA",
page_token = cafes_dorchester_next1$next_page_token,
key = api)
cafes_charles <- google_places(search_string = "Coffee roaster in Charlestown, MA",
key = api)
cafes_charles_next1 <- google_places(search_string = "Coffee roaster in Charlestown, MA",
page_token = cafes_charles$next_page_token,
key = api)
cafes_charles_next2 <- google_places(search_string = "Coffee roaster in Charlestown, MA",
page_token = cafes_charles_next1$next_page_token,
key = api)
cafes_brighton <- google_places(search_string = "Coffee roaster in Brighton, MA",
key = api)
cafes_brighton_next1 <- google_places(search_string = "Coffee roaster in Brighton, MA",
page_token = cafes_brighton$next_page_token,
key = api)
cafes_brighton_next2 <- google_places(search_string = "Coffee roaster in Brighton, MA",
page_token = cafes_brighton_next1$next_page_token,
key = api)
cafes_newton <- google_places(search_string = "Coffee roaster in Newton, MA",
key = api)
cafes_newton_next1 <- google_places(search_string = "Coffee roaster in Newton, MA",
page_token = cafes_newton$next_page_token,
key = api)
cafes_newton_next2 <- google_places(search_string = "Coffee roaster in Newton, MA",
page_token = cafes_newton_next1$next_page_token,
key = api)
cafes_merged <- cafes_somerville$results %>%
as_tibble() %>%
bind_rows(cafes_cambridge$results) %>%
bind_rows(cafes_boston$results) %>%
bind_rows(cafes_brookline$results) %>%
bind_rows(cafes_jp$results) %>%
bind_rows(cafes_medford$results) %>%
bind_rows(cafes_belmont$results) %>%
bind_rows(cafes_dorchester$results) %>%
bind_rows(cafes_charles$results) %>%
bind_rows(cafes_brighton$results) %>%
bind_rows(cafes_newton$results) %>%
bind_rows(cafes_somerville_next1$results) %>%
bind_rows(cafes_somerville_next2$results) %>%
bind_rows(cafes_cambridge_next1$results) %>%
bind_rows(cafes_cambridge_next2$results) %>%
bind_rows(cafes_boston_next1$results) %>%
bind_rows(cafes_boston_next2$results) %>%
bind_rows(cafes_brookline_next1$results) %>%
bind_rows(cafes_brookline_next2$results) %>%
bind_rows(cafes_jp_next1$results) %>%
bind_rows(cafes_jp_next2$results) %>%
bind_rows(cafes_medford_next1$results) %>%
bind_rows(cafes_medford_next2$results) %>%
bind_rows(cafes_belmont_next1$results) %>%
bind_rows(cafes_belmont_next2$results) %>%
bind_rows(cafes_dorchester_next1$results) %>%
bind_rows(cafes_dorchester_next2$results) %>%
bind_rows(cafes_charles_next1$results) %>%
bind_rows(cafes_charles_next2$results) %>%
bind_rows(cafes_brighton_next1$results) %>%
bind_rows(cafes_brighton_next2$results) %>%
bind_rows(cafes_newton_next1$results) %>%
bind_rows(cafes_newton_next2$results) %>%
unnest(cols = geometry) %>%
unnest(location) %>%
dplyr::select(name,lat,lng,rating,user_ratings_total) %>%
dplyr::filter(rating > 4.25 & user_ratings_total > 10) %>%
dplyr::distinct(name, lat, lng, .keep_all=TRUE)Peek at result, then buffer by max walking distance
cafes_merged# A tibble: 110 × 5
name lat lng rating user_ratings_total
<chr> <dbl> <dbl> <dbl> <int>
1 Yego Coffee 42.4 -71.1 5 213
2 Zaruma Gold Coffee 42.4 -71.1 4.5 208
3 Wicked Cafe 42.4 -71.1 4.8 58
4 Broadsheet Coffee Roasters 42.4 -71.1 4.6 613
5 True Grounds 42.4 -71.1 4.5 322
6 Nine Bar Espresso 42.4 -71.1 4.6 335
7 Maprang Bakery & Cafe 42.4 -71.1 4.9 156
8 Bloc Cafe 42.4 -71.1 4.4 765
9 Diesel Cafe 42.4 -71.1 4.4 1277
10 3 Little Figs 42.4 -71.1 4.7 954
# ℹ 100 more rowscafes_merged_buffered <- st_as_sf(cafes_merged, coords = c("lng","lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters)Find which census blocks have a cafe in walking distance
blocks_walkToCafe <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,cafes_merged_buffered)Summarize number of cafes found within walking distance per block, and create a binned score based on these counts
nCafes_perBlock <- lengths(blocks_walkToCafe) %>%
as_tibble() %>%
mutate(nCafe = case_when(
value == 0 ~ 0,
(value > 0) & (value < 3) ~ 1,
value >= 3 ~ 2
)
)
greaterBoston_censusblocks$nCafes <- nCafes_perBlock$nCafePlot base map then color census blocks by cafe score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(greaterBoston_censusblocks,crs = 4326),
fill_colour = "nCafes",
fill_opacity = 0.75,
legend = T) 
Retrieve listings for highly-reviewed restaurants in the greater Boston area
restaurants_somerville <- google_places(search_string = "Restaurant in Somerville, MA",
key = api)
restaurants_somerville_next1 <- google_places(search_string = "Restaurant in Somerville, MA",
page_token = restaurants_somerville$next_page_token,
key = api)
restaurants_somerville_next2 <- google_places(search_string = "Restaurant in Somerville, MA",
page_token = restaurants_somerville_next1$next_page_token,
key = api)
restaurants_cambridge <- google_places(search_string = "Restaurant in Cambridge, MA",
key = api)
restaurants_cambridge_next1 <- google_places(search_string = "Restaurant in Cambridge, MA",
page_token = restaurants_cambridge$next_page_token,
key = api)
restaurants_cambridge_next2 <- google_places(search_string = "Restaurant in Cambridge, MA",
page_token = restaurants_cambridge_next1$next_page_token,
key = api)
restaurants_boston <- google_places(search_string = "Restaurant in Boston, MA",
key = api)
restaurants_boston_next1 <- google_places(search_string = "Restaurant in Boston, MA",
page_token = restaurants_boston$next_page_token,
key = api)
restaurants_boston_next2 <- google_places(search_string = "Restaurant in Boston, MA",
page_token = restaurants_boston_next1$next_page_token,
key = api)
restaurants_brookline <- google_places(search_string = "Restaurant in Brookline, MA",
key = api)
restaurants_brookline_next1 <- google_places(search_string = "Restaurant in Brookline, MA",
page_token = restaurants_brookline$next_page_token,
key = api)
restaurants_brookline_next2 <- google_places(search_string = "Restaurant in Brookline, MA",
page_token = restaurants_brookline_next1$next_page_token,
key = api)
restaurants_jp <- google_places(search_string = "Restaurant in Jamaica Plain, MA",
key = api)
restaurants_jp_next1 <- google_places(search_string = "Restaurant in Jamaica Plain, MA",
page_token = restaurants_jp$next_page_token,
key = api)
restaurants_jp_next2 <- google_places(search_string = "Restaurant in Jamaica Plain, MA",
page_token = restaurants_jp_next1$next_page_token,
key = api)
restaurants_medford <- google_places(search_string = "Restaurant in Medford, MA",
key = api)
restaurants_medford_next1 <- google_places(search_string = "Restaurant in Medford, MA",
page_token = restaurants_medford$next_page_token,
key = api)
restaurants_medford_next2 <- google_places(search_string = "Restaurant in Medford, MA",
page_token = restaurants_medford_next1$next_page_token,
key = api)
restaurants_belmont <- google_places(search_string = "Restaurant in Belmont, MA",
key = api)
restaurants_belmont_next1 <- google_places(search_string = "Restaurant in Belmont, MA",
page_token = restaurants_belmont$next_page_token,
key = api)
restaurants_belmont_next2 <- google_places(search_string = "Restaurant in Belmont, MA",
page_token = restaurants_belmont_next1$next_page_token,
key = api)
restaurants_dorchester <- google_places(search_string = "Restaurant in Dorchester, MA",
key = api)
restaurants_dorchester_next1 <- google_places(search_string = "Restaurant in Dorchester, MA",
page_token = restaurants_dorchester$next_page_token,
key = api)
restaurants_dorchester_next2 <- google_places(search_string = "Restaurant in Dorchester, MA",
page_token = restaurants_dorchester_next1$next_page_token,
key = api)
restaurants_charles <- google_places(search_string = "Restaurant in Charlestown, MA",
key = api)
restaurants_charles_next1 <- google_places(search_string = "Restaurant in Charlestown, MA",
page_token = restaurants_charles$next_page_token,
key = api)
restaurants_charles_next2 <- google_places(search_string = "Restaurant in Charlestown, MA",
page_token = restaurants_charles_next1$next_page_token,
key = api)
restaurants_brighton <- google_places(search_string = "Restaurant in Brighton, MA",
key = api)
restaurants_brighton_next1 <- google_places(search_string = "Restaurant in Brighton, MA",
page_token = restaurants_brighton$next_page_token,
key = api)
restaurants_brighton_next2 <- google_places(search_string = "Restaurant in Brighton, MA",
page_token = restaurants_brighton_next1$next_page_token,
key = api)
restaurants_newton <- google_places(search_string = "Restaurant in Newton, MA",
key = api)
restaurants_newton_next1 <- google_places(search_string = "Restaurant in Newton, MA",
page_token = restaurants_newton$next_page_token,
key = api)
restaurants_newton_next2 <- google_places(search_string = "Restaurant in Newton, MA",
page_token = restaurants_newton_next1$next_page_token,
key = api)
restaurants_merged <- restaurants_somerville$results %>%
as_tibble() %>%
bind_rows(restaurants_cambridge$results) %>%
bind_rows(restaurants_boston$results) %>%
bind_rows(restaurants_brookline$results) %>%
bind_rows(restaurants_jp$results) %>%
bind_rows(restaurants_medford$results) %>%
bind_rows(restaurants_belmont$results) %>%
bind_rows(restaurants_dorchester$results) %>%
bind_rows(restaurants_charles$results) %>%
bind_rows(restaurants_brighton$results) %>%
bind_rows(restaurants_newton$results) %>%
bind_rows(restaurants_somerville_next1$results) %>%
bind_rows(restaurants_somerville_next2$results) %>%
bind_rows(restaurants_cambridge_next1$results) %>%
bind_rows(restaurants_cambridge_next2$results) %>%
bind_rows(restaurants_boston_next1$results) %>%
bind_rows(restaurants_boston_next2$results) %>%
bind_rows(restaurants_brookline_next1$results) %>%
bind_rows(restaurants_brookline_next2$results) %>%
bind_rows(restaurants_jp_next1$results) %>%
bind_rows(restaurants_jp_next2$results) %>%
bind_rows(restaurants_medford_next1$results) %>%
bind_rows(restaurants_medford_next2$results) %>%
bind_rows(restaurants_belmont_next1$results) %>%
bind_rows(restaurants_belmont_next2$results) %>%
bind_rows(restaurants_dorchester_next1$results) %>%
bind_rows(restaurants_dorchester_next2$results) %>%
bind_rows(restaurants_charles_next1$results) %>%
bind_rows(restaurants_charles_next2$results) %>%
bind_rows(restaurants_brighton_next1$results) %>%
bind_rows(restaurants_brighton_next2$results) %>%
bind_rows(restaurants_newton_next1$results) %>%
bind_rows(restaurants_newton_next2$results) %>%
unnest(cols = geometry) %>%
unnest(location) %>%
dplyr::select(name,lat,lng,rating,user_ratings_total) %>%
dplyr::filter(rating > 4 & user_ratings_total > 10) %>%
dplyr::distinct(name, lat, lng, .keep_all=TRUE)Peek at result, then buffer by max walking distance
restaurants_merged# A tibble: 224 × 5
name lat lng rating user_ratings_total
<chr> <dbl> <dbl> <dbl> <int>
1 The Neighborhood Restaurant 42.4 -71.1 4.7 1435
2 Salt + Stone 42.4 -71.1 4.1 408
3 Hot Tomatoes 42.4 -71.1 4.8 56
4 Civility Social House 42.4 -71.1 4.7 139
5 Highland Kitchen 42.4 -71.1 4.6 1313
6 Josephine 42.4 -71.1 4.2 175
7 Sarma Restaurant 42.4 -71.1 4.7 2172
8 Cristian Latin 42.4 -71.1 4.9 53
9 Trina's Starlite Lounge 42.4 -71.1 4.7 951
10 Juliet 42.4 -71.1 4.5 529
# ℹ 214 more rowsrestaurants_merged_buffered <- st_as_sf(restaurants_merged, coords = c("lng","lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters)Find which census blocks have a restaurant in walking distance
blocks_walkToRestaurant <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,restaurants_merged_buffered)Summarize number of restaurants found within walking distance per block, and create score
nRestaurants_perBlock <- lengths(blocks_walkToRestaurant) %>%
as_tibble() %>%
mutate(nRestaurants = case_when(
value == 0 ~ 0,
(value > 0) & (value < 3) ~ 1,
value >= 3 ~ 2
)
)
greaterBoston_censusblocks$nRestaurants <- nRestaurants_perBlock$nRestaurantsPlot base map then color census blocks by restaurant score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(greaterBoston_censusblocks,crs = 4326),
fill_colour = "nRestaurants",
fill_opacity = 0.75,
legend = T) 
Retrieve listings for highly-reviewed grocery stores in the greater Boston area
grocery_somerville <- google_places(search_string = "Grocery store in Somerville, MA",
key = api)
grocery_somerville_next1 <- google_places(search_string = "Grocery store in Somerville, MA",
page_token = grocery_somerville$next_page_token,
key = api)
grocery_somerville_next2 <- google_places(search_string = "Grocery store in Somerville, MA",
page_token = grocery_somerville_next1$next_page_token,
key = api)
grocery_cambridge <- google_places(search_string = "Grocery store in Cambridge, MA",
key = api)
grocery_cambridge_next1 <- google_places(search_string = "Grocery store in Cambridge, MA",
page_token = grocery_cambridge$next_page_token,
key = api)
grocery_cambridge_next2 <- google_places(search_string = "Grocery store in Cambridge, MA",
page_token = grocery_cambridge_next1$next_page_token,
key = api)
grocery_boston <- google_places(search_string = "Grocery store in Boston, MA",
key = api)
grocery_boston_next1 <- google_places(search_string = "Grocery store in Boston, MA",
page_token = grocery_boston$next_page_token,
key = api)
grocery_boston_next2 <- google_places(search_string = "Grocery store in Boston, MA",
page_token = grocery_boston_next1$next_page_token,
key = api)
grocery_brookline <- google_places(search_string = "Grocery store in Brookline, MA",
key = api)
grocery_brookline_next1 <- google_places(search_string = "Grocery store in Brookline, MA",
page_token = grocery_brookline$next_page_token,
key = api)
grocery_brookline_next2 <- google_places(search_string = "Grocery store in Brookline, MA",
page_token = grocery_brookline_next1$next_page_token,
key = api)
grocery_jp <- google_places(search_string = "Grocery store in Jamaica Plain, MA",
key = api)
grocery_jp_next1 <- google_places(search_string = "Grocery store in Jamaica Plain, MA",
page_token = grocery_jp$next_page_token,
key = api)
grocery_jp_next2 <- google_places(search_string = "Grocery store in Jamaica Plain, MA",
page_token = grocery_jp_next1$next_page_token,
key = api)
grocery_medford <- google_places(search_string = "Grocery store in Medford, MA",
key = api)
grocery_medford_next1 <- google_places(search_string = "Grocery store in Medford, MA",
page_token = grocery_medford$next_page_token,
key = api)
grocery_medford_next2 <- google_places(search_string = "Grocery store in Medford, MA",
page_token = grocery_medford_next1$next_page_token,
key = api)
grocery_belmont <- google_places(search_string = "Grocery store in Belmont, MA",
key = api)
grocery_belmont_next1 <- google_places(search_string = "Grocery store in Belmont, MA",
page_token = grocery_belmont$next_page_token,
key = api)
grocery_belmont_next2 <- google_places(search_string = "Grocery store in Belmont, MA",
page_token = grocery_belmont_next1$next_page_token,
key = api)
grocery_dorchester <- google_places(search_string = "Grocery store in Dorchester, MA",
key = api)
grocery_dorchester_next1 <- google_places(search_string = "Grocery store in Dorchester, MA",
page_token = grocery_dorchester$next_page_token,
key = api)
grocery_dorchester_next2 <- google_places(search_string = "Grocery store in Dorchester, MA",
page_token = grocery_dorchester_next1$next_page_token,
key = api)
grocery_charles <- google_places(search_string = "Grocery store in Charlestown, MA",
key = api)
grocery_charles_next1 <- google_places(search_string = "Grocery store in Charlestown, MA",
page_token = grocery_charles$next_page_token,
key = api)
grocery_charles_next2 <- google_places(search_string = "Grocery store in Charlestown, MA",
page_token = grocery_charles_next1$next_page_token,
key = api)
grocery_brighton <- google_places(search_string = "Grocery store in Brighton, MA",
key = api)
grocery_brighton_next1 <- google_places(search_string = "Grocery store in Brighton, MA",
page_token = grocery_brighton$next_page_token,
key = api)
grocery_brighton_next2 <- google_places(search_string = "Grocery store in Brighton, MA",
page_token = grocery_brighton_next1$next_page_token,
key = api)
grocery_newton <- google_places(search_string = "Grocery store in Newton, MA",
key = api)
grocery_newton_next1 <- google_places(search_string = "Grocery store in Newton, MA",
page_token = grocery_newton$next_page_token,
key = api)
grocery_newton_next2 <- google_places(search_string = "Grocery store in Newton, MA",
page_token = grocery_newton_next1$next_page_token,
key = api)
grocery_merged <- grocery_somerville$results %>%
as_tibble() %>%
bind_rows(grocery_cambridge$results) %>%
bind_rows(grocery_boston$results) %>%
bind_rows(grocery_brookline$results) %>%
bind_rows(grocery_jp$results) %>%
bind_rows(grocery_medford$results) %>%
bind_rows(grocery_belmont$results) %>%
bind_rows(grocery_dorchester$results) %>%
bind_rows(grocery_charles$results) %>%
bind_rows(grocery_brighton$results) %>%
bind_rows(grocery_newton$results) %>%
bind_rows(grocery_somerville_next1$results) %>%
bind_rows(grocery_somerville_next2$results) %>%
bind_rows(grocery_cambridge_next1$results) %>%
bind_rows(grocery_cambridge_next2$results) %>%
bind_rows(grocery_boston_next1$results) %>%
bind_rows(grocery_boston_next2$results) %>%
bind_rows(grocery_brookline_next1$results) %>%
bind_rows(grocery_brookline_next2$results) %>%
bind_rows(grocery_jp_next1$results) %>%
bind_rows(grocery_jp_next2$results) %>%
bind_rows(grocery_medford_next1$results) %>%
bind_rows(grocery_medford_next2$results) %>%
bind_rows(grocery_belmont_next1$results) %>%
bind_rows(grocery_belmont_next2$results) %>%
bind_rows(grocery_dorchester_next1$results) %>%
bind_rows(grocery_dorchester_next2$results) %>%
bind_rows(grocery_charles_next1$results) %>%
bind_rows(grocery_charles_next2$results) %>%
bind_rows(grocery_brighton_next1$results) %>%
bind_rows(grocery_brighton_next2$results) %>%
bind_rows(grocery_newton_next1$results) %>%
bind_rows(grocery_newton_next2$results) %>%
unnest(cols = geometry) %>%
unnest(location) %>%
dplyr::select(name,lat,lng,rating,user_ratings_total) %>%
dplyr::filter(rating > 4 & user_ratings_total > 10) %>%
dplyr::distinct(name, lat, lng, .keep_all=TRUE)Peek at result, then buffer by max walking distance
grocery_merged# A tibble: 107 × 5
name lat lng rating user_ratings_total
<chr> <dbl> <dbl> <dbl> <int>
1 Market Basket 42.4 -71.1 4.4 4433
2 Stop & Shop 42.4 -71.1 4.2 1362
3 Whole Foods Market 42.4 -71.1 4.3 764
4 Neighborhood Produce 42.4 -71.1 4.9 153
5 Wegmans 42.4 -71.1 4.5 3005
6 Get-N-Go 42.4 -71.1 4.9 25
7 Broadway Marketplace 42.4 -71.1 4.1 462
8 Shree Bombay Market 42.4 -71.1 4.3 113
9 Whole Foods Market 42.4 -71.1 4.1 863
10 Reliable Market 42.4 -71.1 4.5 581
# ℹ 97 more rowsgrocery_merged_buffered <- st_as_sf(grocery_merged, coords = c("lng","lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters)Find which census blocks have a grocery store in walking distance
blocks_walkToGrocery <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,grocery_merged_buffered)Summarize number of grocery stores found within walking distance per block, and create score
nGrocery_perBlock <- lengths(blocks_walkToGrocery) %>%
as_tibble() %>%
mutate(nGrocery = case_when(
value == 0 ~ 0,
(value > 0) & (value < 3) ~ 1,
value >= 3 ~ 2
)
)
greaterBoston_censusblocks$nGrocery <- nGrocery_perBlock$nGroceryPlot base map then color census blocks by Grocery score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(greaterBoston_censusblocks,crs = 4326),
fill_colour = "nGrocery",
fill_opacity = 0.75,
legend = T) 
Retrieve listings for highly-reviewed parks and playgrounds in the greater Boston area
parks_somerville <- google_places(search_string = "Parks and playgrounds in Somerville, MA",
key = api)
parks_somerville_next1 <- google_places(search_string = "Parks and playgrounds in Somerville, MA",
page_token = parks_somerville$next_page_token,
key = api)
parks_somerville_next2 <- google_places(search_string = "Parks and playgrounds in Somerville, MA",
page_token = parks_somerville_next1$next_page_token,
key = api)
parks_cambridge <- google_places(search_string = "Parks and playgrounds in Cambridge, MA",
key = api)
parks_cambridge_next1 <- google_places(search_string = "Parks and playgrounds in Cambridge, MA",
page_token = parks_cambridge$next_page_token,
key = api)
parks_cambridge_next2 <- google_places(search_string = "Parks and playgrounds in Cambridge, MA",
page_token = parks_cambridge_next1$next_page_token,
key = api)
parks_boston <- google_places(search_string = "Parks and playgrounds in Boston, MA",
key = api)
parks_boston_next1 <- google_places(search_string = "Parks and playgrounds in Boston, MA",
page_token = parks_boston$next_page_token,
key = api)
parks_boston_next2 <- google_places(search_string = "Parks and playgrounds in Boston, MA",
page_token = parks_boston_next1$next_page_token,
key = api)
parks_brookline <- google_places(search_string = "Parks and playgrounds in Brookline, MA",
key = api)
parks_brookline_next1 <- google_places(search_string = "Parks and playgrounds in Brookline, MA",
page_token = parks_brookline$next_page_token,
key = api)
parks_brookline_next2 <- google_places(search_string = "Parks and playgrounds in Brookline, MA",
page_token = parks_brookline_next1$next_page_token,
key = api)
parks_jp <- google_places(search_string = "Parks and playgrounds in Jamaica Plain, MA",
key = api)
parks_jp_next1 <- google_places(search_string = "Parks and playgrounds in Jamaica Plain, MA",
page_token = parks_jp$next_page_token,
key = api)
parks_jp_next2 <- google_places(search_string = "Parks and playgrounds in Jamaica Plain, MA",
page_token = parks_jp_next1$next_page_token,
key = api)
parks_medford <- google_places(search_string = "Parks and playgrounds in Medford, MA",
key = api)
parks_medford_next1 <- google_places(search_string = "Parks and playgrounds in Medford, MA",
page_token = parks_medford$next_page_token,
key = api)
parks_medford_next2 <- google_places(search_string = "Parks and playgrounds in Medford, MA",
page_token = parks_medford_next1$next_page_token,
key = api)
parks_belmont <- google_places(search_string = "Parks and playgrounds in Belmont, MA",
key = api)
parks_belmont_next1 <- google_places(search_string = "Parks and playgrounds in Belmont, MA",
page_token = parks_belmont$next_page_token,
key = api)
parks_belmont_next2 <- google_places(search_string = "Parks and playgrounds in Belmont, MA",
page_token = parks_belmont_next1$next_page_token,
key = api)
parks_dorchester <- google_places(search_string = "Parks and playgrounds in Dorchester, MA",
key = api)
parks_dorchester_next1 <- google_places(search_string = "Parks and playgrounds in Dorchester, MA",
page_token = parks_dorchester$next_page_token,
key = api)
parks_dorchester_next2 <- google_places(search_string = "Parks and playgrounds in Dorchester, MA",
page_token = parks_dorchester_next1$next_page_token,
key = api)
parks_charles <- google_places(search_string = "Parks and playgrounds in Charlestown, MA",
key = api)
parks_charles_next1 <- google_places(search_string = "Parks and playgrounds in Charlestown, MA",
page_token = parks_charles$next_page_token,
key = api)
parks_charles_next2 <- google_places(search_string = "Parks and playgrounds in Charlestown, MA",
page_token = parks_charles_next1$next_page_token,
key = api)
parks_brighton <- google_places(search_string = "Parks and playgrounds in Brighton, MA",
key = api)
parks_brighton_next1 <- google_places(search_string = "Parks and playgrounds in Brighton, MA",
page_token = parks_brighton$next_page_token,
key = api)
parks_brighton_next2 <- google_places(search_string = "Parks and playgrounds in Brighton, MA",
page_token = parks_brighton_next1$next_page_token,
key = api)
parks_newton <- google_places(search_string = "Parks and playgrounds in Newton, MA",
key = api)
parks_newton_next1 <- google_places(search_string = "Parks and playgrounds in Newton, MA",
page_token = parks_newton$next_page_token,
key = api)
parks_newton_next2 <- google_places(search_string = "Parks and playgrounds in Newton, MA",
page_token = parks_newton_next1$next_page_token,
key = api)
parks_merged <- parks_somerville$results %>%
as_tibble() %>%
bind_rows(parks_cambridge$results) %>%
bind_rows(parks_boston$results) %>%
bind_rows(parks_brookline$results) %>%
bind_rows(parks_jp$results) %>%
bind_rows(parks_medford$results) %>%
bind_rows(parks_belmont$results) %>%
bind_rows(parks_dorchester$results) %>%
bind_rows(parks_charles$results) %>%
bind_rows(parks_brighton$results) %>%
bind_rows(parks_newton$results) %>%
bind_rows(parks_somerville_next1$results) %>%
bind_rows(parks_somerville_next2$results) %>%
bind_rows(parks_cambridge_next1$results) %>%
bind_rows(parks_cambridge_next2$results) %>%
bind_rows(parks_boston_next1$results) %>%
bind_rows(parks_boston_next2$results) %>%
bind_rows(parks_brookline_next1$results) %>%
bind_rows(parks_brookline_next2$results) %>%
bind_rows(parks_jp_next1$results) %>%
bind_rows(parks_jp_next2$results) %>%
bind_rows(parks_medford_next1$results) %>%
bind_rows(parks_medford_next2$results) %>%
bind_rows(parks_belmont_next1$results) %>%
bind_rows(parks_belmont_next2$results) %>%
bind_rows(parks_dorchester_next1$results) %>%
bind_rows(parks_dorchester_next2$results) %>%
bind_rows(parks_charles_next1$results) %>%
bind_rows(parks_charles_next2$results) %>%
bind_rows(parks_brighton_next1$results) %>%
bind_rows(parks_brighton_next2$results) %>%
bind_rows(parks_newton_next1$results) %>%
bind_rows(parks_newton_next2$results) %>%
unnest(cols = geometry) %>%
unnest(location) %>%
dplyr::select(name,lat,lng,rating,user_ratings_total) %>%
dplyr::filter(rating > 4 & user_ratings_total > 10) %>%
dplyr::distinct(name, lat, lng, .keep_all=TRUE)Peek at result, then buffer by max walking distance
parks_merged# A tibble: 167 × 5
name lat lng rating user_ratings_total
<chr> <dbl> <dbl> <dbl> <int>
1 Sylvester Baxter Riverfront Park 42.4 -71.1 4.6 540
2 Hoyt-Sullivan Playground 42.4 -71.1 4.7 92
3 Dickerman Playground 42.4 -71.1 4.7 47
4 Lincoln Park 42.4 -71.1 4.8 547
5 Seven Hills Park 42.4 -71.1 4.2 99
6 Morse-Kelley Playground 42.4 -71.1 4.6 35
7 Glen Park 42.4 -71.1 4.6 48
8 Conway Park 42.4 -71.1 4.4 124
9 Albion Playground 42.4 -71.1 4.6 116
10 Prospect Hill Park 42.4 -71.1 4.6 337
# ℹ 157 more rowsparks_merged_buffered <- st_as_sf(parks_merged, coords = c("lng","lat"),crs = 4326) %>%
sf::st_buffer(max_walk_meters)Find which census blocks have a park or playground in walking distance
blocks_walkToParks <- greaterBoston_censusblocks %>%
sf::st_transform(crs = 4326) %>%
sf::st_intersects(.,parks_merged_buffered)Summarize number of parks or playgrounds found within walking distance per block, and create score
nParks_perBlock <- lengths(blocks_walkToParks) %>%
as_tibble() %>%
mutate(nParks = case_when(
value == 0 ~ 0,
(value > 0) & (value < 3) ~ 1,
value >= 3 ~ 2
)
)
greaterBoston_censusblocks$nParks <- nParks_perBlock$nParksPlot base map then color census blocks by parks score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(greaterBoston_censusblocks,crs = 4326),
fill_colour = "nParks",
fill_opacity = 0.75,
legend = T) 
Compute overall walkability score
This is the sum of the T connectivity score above, as well as the density scores we calculated for cafes, restaurants, groceries, and parks within reasonable walking distance. This intentionally weights the T-connectivity score slightly, since accessibility via public transport is important to us.
walkabilityScore <- cbind("T_connectivity" = greaterBoston_censusblocks$T_connectivity,
"nCafes" = greaterBoston_censusblocks$nCafes,
"nRestaurants" = greaterBoston_censusblocks$nRestaurants,
"nGrocery" = greaterBoston_censusblocks$nGrocery,
"nParks" = greaterBoston_censusblocks$nParks
) %>%
as_tibble() %>%
rowwise() %>%
mutate(walkability = sum(T_connectivity,nCafes,nRestaurants,nGrocery,nParks))
walkabilityScore# A tibble: 14,449 × 6
# Rowwise:
T_connectivity nCafes nRestaurants nGrocery nParks walkability
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 1 0 0 0 0 1
2 1 1 0 1 0 3
3 2 0 0 1 2 5
4 2 0 0 1 1 4
5 1 0 0 0 0 1
6 2 2 2 2 2 10
7 2 0 0 0 0 2
8 2 0 0 0 0 2
9 1 0 0 0 0 1
10 1 0 0 0 0 1
# ℹ 14,439 more rowsgreaterBoston_censusblocks$walkabilityScore <- walkabilityScore$walkabilityPlot base map then color census blocks by walkability score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(greaterBoston_censusblocks,crs = 4326),
fill_colour = "walkabilityScore",
fill_opacity = 0.75,
legend = T) 
Compute commute time from the centroid of filtered census blocks to each of our workplaces in the Longwood Medical Area and MGH
Set destination locations and example departure time
work1 <- "3 Blackfan Street, Boston, MA"
work2 <- "55 Fruit St, Boston, MA"
monday_morning <- as.POSIXct("2024-07-08 09:00:00", tz = "America/New_York")Compute centroid of each census block
Optionally, first filter our census blocks by various criteria to cut down on number of API calls. Doing this many times can create a large bill, so plan and filter accordingly. Here I require a minimum walkability score of 9 to proceed
census_centroids <- greaterBoston_censusblocks[greaterBoston_censusblocks$walkabilityScore>=9,] %>%
sf::st_transform(crs = 4326) %>%
sf::st_centroid() %>%
sf::st_coordinates()Warning: st_centroid assumes attributes are constant over geometriesCompute these in chunks of 25 or fewer, splitting into 200 bins makes the max group size <25 here
census_centroids_chunks <- as.data.frame(census_centroids) %>%
as_tibble() %>%
mutate(Chunk = ntile(n=200)) %>%
dplyr::select(Y,X,Chunk) %>%
group_split(Chunk,.keep=F)
commuteTimes <- list()
maxCommutes <- list()
for(i in 1:length(census_centroids_chunks)) {
commuteTimes[[i]] <- googleway::google_distance(
origins = census_centroids_chunks[[i]],
destinations = c(work1, work2),
mode = "transit",
arrival_time = monday_morning,
units = "imperial",
transit_routing_preference = "fewer_transfers",
traffic_model = "best_guess",
key = api
)
maxCommutes[[i]] <- enframe(commuteTimes[[i]]$rows$elements) %>%
unnest(cols=value) %>%
unnest(cols=duration) %>%
dplyr::select(name,value) %>%
mutate(CommuteTime = value/60) %>%
group_by(name) %>%
summarize(MaxCommute = max(CommuteTime))
}Create a binned commute score where the shortest max time in minutes is highest
maxCommutesFlat <- enframe(maxCommutes) %>%
dplyr::rename("Chunk" = name) %>%
unnest(cols=c(value)) %>%
mutate(MaxCommute = round(MaxCommute)) %>%
mutate(CommuteScore = case_when(
MaxCommute >= 55 ~ 0,
MaxCommute >= 45 & MaxCommute < 55 ~ 1,
MaxCommute >= 35 & MaxCommute < 45 ~ 2,
MaxCommute >= 25 & MaxCommute < 35 ~ 3,
MaxCommute < 25 ~ 4
)
)
maxCommutesFlat# A tibble: 3,519 × 4
Chunk name MaxCommute CommuteScore
<int> <int> <dbl> <dbl>
1 1 1 47 1
2 1 2 24 4
3 1 3 56 0
4 1 4 55 0
5 1 5 38 2
6 1 6 31 3
7 1 7 56 0
8 1 8 30 3
9 1 9 42 2
10 1 10 29 3
# ℹ 3,509 more rows# Add these scores to our census blocks data matrix
filtered_blocks <- greaterBoston_censusblocks[greaterBoston_censusblocks$walkabilityScore>=9,]
filtered_blocks <- filtered_blocks[filtered_blocks$POP20 > 0,]
filtered_blocks$CommuteScore <- maxCommutesFlat$CommuteScorePlot base map then color census blocks by commute score
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(filtered_blocks,crs = 4326),
fill_colour = "CommuteScore",
fill_opacity = 0.75,
legend = T) 
Retrieve MLS listings for each greater Boston town using HomeHarvest
This is a python package that is easy to set up and run; it pulls listings from Realtor.com and we create a csv tabulating each Boston-area town’s for-sale listings within the last 30 days
from homeharvest import scrape_property
from datetime import datetime
current_timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
Arlington_filename = f"HomeHarvest_Arlington_{current_timestamp}.csv"
Arlington_properties = scrape_property(
location="Arlington, MA",
listing_type="for_sale",
past_days=30
)
Arlington_properties.to_csv(Arlington_filename, index=False)
Belmont_filename = f"HomeHarvest_Belmont_{current_timestamp}.csv"
Belmont_properties = scrape_property(
location="Belmont, MA",
listing_type="for_sale",
past_days=30
)
Belmont_properties.to_csv(Belmont_filename, index=False)
Boston_filename = f"HomeHarvest_Boston_{current_timestamp}.csv"
Boston_properties = scrape_property(
location="Boston, MA",
listing_type="for_sale",
past_days=30
)
Boston_properties.to_csv(Boston_filename, index=False)
Braintree_filename = f"HomeHarvest_Braintree_{current_timestamp}.csv"
Braintree_properties = scrape_property(
location="Braintree, MA",
listing_type="for_sale",
past_days=30
)
Braintree_properties.to_csv(Braintree_filename, index=False)
Brookline_filename = f"HomeHarvest_Brookline_{current_timestamp}.csv"
Brookline_properties = scrape_property(
location="Brookline, MA",
listing_type="for_sale",
past_days=30
)
Brookline_properties.to_csv(Brookline_filename, index=False)
Cambridge_filename = f"HomeHarvest_Cambridge_{current_timestamp}.csv"
Cambridge_properties = scrape_property(
location="Cambridge, MA",
listing_type="for_sale",
past_days=30
)
Cambridge_properties.to_csv(Cambridge_filename, index=False)
Chelsea_filename = f"HomeHarvest_Chelsea_{current_timestamp}.csv"
Chelsea_properties = scrape_property(
location="Chelsea, MA",
listing_type="for_sale",
past_days=30
)
Chelsea_properties.to_csv(Chelsea_filename, index=False)
Everett_filename = f"HomeHarvest_Everett_{current_timestamp}.csv"
Everett_properties = scrape_property(
location="Everett, MA",
listing_type="for_sale",
past_days=30
)
Everett_properties.to_csv(Everett_filename, index=False)
Malden_filename = f"HomeHarvest_Malden_{current_timestamp}.csv"
Malden_properties = scrape_property(
location="Malden, MA",
listing_type="for_sale",
past_days=30
)
Malden_properties.to_csv(Malden_filename, index=False)
Medford_filename = f"HomeHarvest_Medford_{current_timestamp}.csv"
Medford_properties = scrape_property(
location="Medford, MA",
listing_type="for_sale",
past_days=30
)
Medford_properties.to_csv(Medford_filename, index=False)
Melrose_filename = f"HomeHarvest_Melrose_{current_timestamp}.csv"
Melrose_properties = scrape_property(
location="Melrose, MA",
listing_type="for_sale",
past_days=30
)
Melrose_properties.to_csv(Melrose_filename, index=False)
Milton_filename = f"HomeHarvest_Milton_{current_timestamp}.csv"
Milton_properties = scrape_property(
location="Milton, MA",
listing_type="for_sale",
past_days=30
)
Milton_properties.to_csv(Milton_filename, index=False)
Newton_filename = f"HomeHarvest_Newton_{current_timestamp}.csv"
Newton_properties = scrape_property(
location="Newton, MA",
listing_type="for_sale",
past_days=30
)
Newton_properties.to_csv(Newton_filename, index=False)
Quincy_filename = f"HomeHarvest_Quincy_{current_timestamp}.csv"
Quincy_properties = scrape_property(
location="Quincy, MA",
listing_type="for_sale",
past_days=30
)
Quincy_properties.to_csv(Quincy_filename, index=False)
Revere_filename = f"HomeHarvest_Revere_{current_timestamp}.csv"
Revere_properties = scrape_property(
location="Revere, MA",
listing_type="for_sale",
past_days=30
)
Revere_properties.to_csv(Revere_filename, index=False)
Somerville_filename = f"HomeHarvest_Somerville_{current_timestamp}.csv"
Somerville_properties = scrape_property(
location="Somerville, MA",
listing_type="for_sale",
past_days=30
)
Somerville_properties.to_csv(Somerville_filename, index=False)
Waltham_filename = f"HomeHarvest_Waltham_{current_timestamp}.csv"
Waltham_properties = scrape_property(
location="Waltham, MA",
listing_type="for_sale",
past_days=30
)
Waltham_properties.to_csv(Waltham_filename, index=False)
Watertown_filename = f"HomeHarvest_Watertown_{current_timestamp}.csv"
Watertown_properties = scrape_property(
location="Watertown, MA",
listing_type="for_sale",
past_days=30
)
Watertown_properties.to_csv(Watertown_filename, index=False)
Winthrop_filename = f"HomeHarvest_Winthrop_{current_timestamp}.csv"
Winthrop_properties = scrape_property(
location="Winthrop, MA",
listing_type="for_sale",
past_days=30
)
Winthrop_properties.to_csv(Winthrop_filename, index=False)Read MLS results back into R, then overlay onto our scored census blocks and plot
Import MLS results
file.list <- list.files(pattern="HomeHarvest_*")
df.list <- lapply(file.list, function(x){read_csv(x) %>% mutate(mls_id = as.character(mls_id))})
mls_df <- bind_rows(df.list, .id = "id")
mls_df# A tibble: 1,021 × 44
id property_url mls mls_id status text style full_street_line street
<chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
1 1 https://www.re… BSMA 73255… PENDI… Enjo… COND… 990 Massachuset… 990 M…
2 1 https://www.re… BSMA 73255… PENDI… Welc… COND… 38 Tanager St U… 38 Ta…
3 1 https://www.re… BSMA 73252… PENDI… East… COND… 57 Wyman Ter Un… 57 Wy…
4 1 https://www.re… BSMA 73259… FOR_S… Inve… MULT… 37 Westminster … 37 We…
5 1 https://www.re… BSMA 73257… FOR_S… Luxu… COND… 17 Mount Vernon… 17 Mo…
6 1 https://www.re… BSMA 73255… FOR_S… Anno… COND… 1025 Massachuse… 1025 …
7 1 https://www.re… BSMA 73255… PENDI… A na… SING… 251 Sylvia St 251 S…
8 1 https://www.re… BSMA 73255… PENDI… The … COND… 5 Colonial Vill… 5 Col…
9 1 https://www.re… BSMA 73255… FOR_S… Anno… COND… 1025 Massachuse… 1025 …
10 1 https://www.re… BSMA 73253… FOR_S… Loca… MULT… 7 Park Ave Ext 7 Par…
# ℹ 1,011 more rows
# ℹ 35 more variables: unit <chr>, city <chr>, state <chr>, zip_code <chr>,
# beds <dbl>, full_baths <dbl>, half_baths <dbl>, sqft <dbl>,
# year_built <dbl>, days_on_mls <dbl>, list_price <dbl>, list_date <date>,
# sold_price <dbl>, last_sold_date <date>, assessed_value <dbl>,
# estimated_value <dbl>, lot_sqft <dbl>, price_per_sqft <dbl>,
# latitude <dbl>, longitude <dbl>, neighborhoods <chr>, county <chr>, …Filter MLS listings
Here we set a filter for 2-4 bedrooms for sale listed under $1.5M as an example query
mls_filtered <- mls_df %>%
dplyr::filter(status == "FOR_SALE") %>%
dplyr::filter(beds >= 2 & beds <=4) %>%
dplyr::filter(list_price <= 1500000) %>%
dplyr::select(mls_id,property_url,full_street_line,city,zip_code,beds,full_baths,half_baths,sqft,list_price,latitude,longitude)
mls_filtered# A tibble: 407 × 12
mls_id property_url full_street_line city zip_code beds full_baths
<chr> <chr> <chr> <chr> <chr> <dbl> <dbl>
1 73257275 https://www.realto… 17 Mount Vernon… Arli… 02476 3 3
2 73255813 https://www.realto… 1025 Massachuse… Arli… 02476 2 2
3 73253422 https://www.realto… 7 Park Ave Ext Arli… 02724 4 2
4 73258441 https://www.realto… 31 Surry Rd Uni… Arli… 02476 4 2
5 73254782 https://www.realto… 56 Wyman St Uni… Arli… 02474 2 1
6 73258370 https://www.realto… 5 Mystic Lake D… Arli… 02474 3 4
7 73254860 https://www.realto… 176 Park Ave Un… Arli… 02476 2 1
8 73258983 https://www.realto… 17 Mount Vernon… Arli… 02478 3 3
9 73258412 https://www.realto… 1025 Massachuse… Arli… 02476 3 2
10 73257584 https://www.realto… 279 Channing Rd Belm… 02478 3 2
# ℹ 397 more rows
# ℹ 5 more variables: half_baths <dbl>, sqft <dbl>, list_price <dbl>,
# latitude <dbl>, longitude <dbl>Find intersection between houses for sale and high-scoring census blocks in greater Boston
Require additionally here that the census block has a high commute score
mls_intersection <- mls_filtered %>%
sf::st_as_sf(coords = c("longitude","latitude"),crs = 4326) %>%
sf::st_intersection(.,st_transform(filtered_blocks[filtered_blocks$CommuteScore>=2,],crs = 4326))Warning: attribute variables are assumed to be spatially constant throughout
all geometriesmls_intersectionSimple feature collection with 31 features and 42 fields
Geometry type: POINT
Dimension: XY
Bounding box: xmin: -71.14921 ymin: 42.31591 xmax: -71.05728 ymax: 42.38341
Geodetic CRS: WGS 84
# A tibble: 31 × 43
mls_id property_url full_street_line city zip_code beds full_baths
* <chr> <chr> <chr> <chr> <chr> <dbl> <dbl>
1 73258788 https://www.realto… 15 Trowbridge S… Camb… 02138 2 1
2 73254345 https://www.realto… 52 Franklin St … Some… 02145 2 2
3 73252063 https://www.realto… 140 Shawmut Ave… Bost… 02118 2 2
4 73257304 https://www.realto… 214 Market St A… Bost… 02135 3 2
5 73250779 https://www.realto… 176 School St U… Bost… 02119 3 2
6 73249889 https://www.realto… 12 Stoneholm St… Bost… 02115 2 2
7 73255101 https://www.realto… 1070 Beacon St … Broo… 02446 2 1
8 73253643 https://www.realto… 18 South St Apt… Some… 02143 2 1
9 73257692 https://www.realto… 108 Peterboroug… Bost… 02215 2 1
10 73253842 https://www.realto… 40 Browne St Ap… Broo… 02446 2 1
# ℹ 21 more rows
# ℹ 36 more variables: half_baths <dbl>, sqft <dbl>, list_price <dbl>,
# STATEFP20 <chr>, COUNTYFP20 <chr>, TRACTCE20 <chr>, BLOCKCE20 <chr>,
# GEOID20 <chr>, NAME20 <chr>, MTFCC20 <chr>, ALAND20 <dbl>, AWATER20 <dbl>,
# INTPTLAT20 <chr>, INTPTLON20 <chr>, HOUSING20 <dbl>, POP20 <dbl>,
# AREA_SQFT <dbl>, AREA_ACRES <dbl>, TOWN <chr>, TOWN_ID <int>,
# BLKGRP20 <chr>, TRACT20 <chr>, COUSUBFP <chr>, SHAPE_AREA <dbl>, …Plot base map showing commute score. Add pins for each listed property
The URL to the underlying MLS listing appears when you click on a pin
google_map(data=t_stops) %>%
add_transit() %>%
add_polygons(data = st_transform(filtered_blocks,crs = 4326),
fill_colour = "CommuteScore",
fill_opacity = 0.75,
legend = T) %>%
add_markers(data=mls_intersection, title="list_price", info_window="property_url")
Prepare a tsv for export showing the intersecting properties and their various computed features for filtering
mls_intersection_forExport <- mls_intersection %>%
as_tibble() %>%
dplyr::select(mls_id,
property_url,
full_street_line,
city,
zip_code,
beds,
full_baths,
half_baths,
sqft,
list_price,
T_connectivity,
nCafes,
nRestaurants,
nGrocery,
nParks,
walkabilityScore,
CommuteScore
)
mls_intersection_forExport# A tibble: 31 × 17
mls_id property_url full_street_line city zip_code beds full_baths
<chr> <chr> <chr> <chr> <chr> <dbl> <dbl>
1 73258788 https://www.realto… 15 Trowbridge S… Camb… 02138 2 1
2 73254345 https://www.realto… 52 Franklin St … Some… 02145 2 2
3 73252063 https://www.realto… 140 Shawmut Ave… Bost… 02118 2 2
4 73257304 https://www.realto… 214 Market St A… Bost… 02135 3 2
5 73250779 https://www.realto… 176 School St U… Bost… 02119 3 2
6 73249889 https://www.realto… 12 Stoneholm St… Bost… 02115 2 2
7 73255101 https://www.realto… 1070 Beacon St … Broo… 02446 2 1
8 73253643 https://www.realto… 18 South St Apt… Some… 02143 2 1
9 73257692 https://www.realto… 108 Peterboroug… Bost… 02215 2 1
10 73253842 https://www.realto… 40 Browne St Ap… Broo… 02446 2 1
# ℹ 21 more rows
# ℹ 10 more variables: half_baths <dbl>, sqft <dbl>, list_price <dbl>,
# T_connectivity <dbl>, nCafes <dbl>, nRestaurants <dbl>, nGrocery <dbl>,
# nParks <dbl>, walkabilityScore <dbl>, CommuteScore <dbl>write_tsv(mls_intersection_forExport,"GoogleMaps_CommuteTimes_CensusCentroids_Walkability9plus_GreaterBoston_OverlappingMLS.tsv")Future ideas
Incorporate non-MBTA transit routes, e.g. Harvard/MGH/MASCO shuttles
Additional filtering criteria per census block
Alternative means for routing directions? Want to both minimize transfers and walking if possible, and have clear transit backup options in case of closure, traffic, etc.