NYC Restaurant Inspection Data

Setup

knitr::opts_chunk$set(fig.width=12, fig.height=8, fig.path='Figs/',
                      warning=FALSE, message=FALSE)

Loading libraries

library(tidyverse)
library(httr)
library(jsonlite)
library(plotly)
library(leaflet)
library(knitr)

theme_set(
  ggthemes::theme_fivethirtyeight() + theme(legend.position = "bottom")
  )

options(
  ggplot2.continuous.colour = "viridis",
  ggplot2.continuous.colour = "viridis"
)

scale_colour_discrete = scale_color_viridis_d
scale_fill_discrete = scale_fill_viridis_d

1. Pulling in the dataset

api_url = "https://data.cityofnewyork.us/resource/43nn-pn8j.csv"

rest_inspect = 
  GET(api_url, query = list("$limit" = 50000)) %>% 
  content("parsed")

2. Data wrangling, cleaning up the dataset.

rest_inspect = rest_inspect %>% 
  filter(grade %in% c("A", "B", "C"), boro != "0")

3. Exploring dataset

Year to Year Trends in New Restaurants

rest_year_plot = 
  rest_inspect %>% 
  filter(str_detect(inspection_type, "Initial")) %>% 
  separate(grade_date, sep="-", into = c("grade_year", "grade_month", "grade_day")) %>%
  count(boro, grade_year) %>% 
  ggplot(aes(x = grade_year, y = n, color = boro)) + 
  geom_col((aes(fill = boro))) +
  labs(
    title = "New Restaurants in New York City",
    x = "Year",
    y = "Number of Restaurants")
rest_year_plot

Grades

Counting the number of restaurants by borough and by grade.

rest_inspect %>% 
  count(boro, grade) %>% 
  pivot_wider(names_from = grade, values_from = n) %>% 
  kable() 

boro	A	B	C
Bronx	1789	383	164
Brooklyn	5007	849	278
Manhattan	7958	1186	506
Queens	4649	759	298
Staten Island	723	112	33

Filtering based on Cuisine

Italian/ Pizza Restaurants in NYC

italian = c("Italian", "Italian/Pizza")
rest_inspect %>% 
  filter(str_detect(cuisine_description, italian)) %>% 
  count(boro, grade) %>% 
  pivot_wider(names_from = grade, values_from = n) %>% 
  kable()

boro	A	B	C
Bronx	52	8	7
Brooklyn	123	15	6
Manhattan	266	43	19
Queens	98	11	3
Staten Island	60	11	4

French Restaurants in NYC

rest_inspect %>% 
  filter(str_detect(cuisine_description, "French")) %>% 
  count(boro, grade) %>% 
  pivot_wider(names_from = grade, values_from = n) %>% 
  kable()

boro	A	B	C
Brooklyn	34	8	3
Manhattan	204	17	6
Queens	15	1	NA

Mexican Restaurants in NYC

rest_inspect %>% 
  filter(str_detect(cuisine_description, "Mexican")) %>% 
  count(boro, grade) %>% 
  pivot_wider(names_from = grade, values_from = n) %>% 
  kable()

boro	A	B	C
Bronx	98	23	14
Brooklyn	209	47	16
Manhattan	255	37	15
Queens	165	28	14
Staten Island	42	10	5

Pan-Asian Restaurants in NYC

asian = c("Asian", "Japanese", "Korean", "Chinese")

rest_grade =
  rest_inspect %>% 
  filter(str_detect(cuisine_description, asian)) %>% 
  count(boro, grade) %>% 
  pivot_wider(names_from = grade, values_from = n) %>% 
  kable()

Restaurant Scores

Boxplot of Scores by Borough

rest_inspect %>% 
  group_by(boro) %>% 
  filter(score %in% 1:30)  %>% 
  ggplot(aes(x = boro, y = score, color = boro)) +
  geom_boxplot() +
  labs(
    title = "Distribution of Scores",
    x = "Boro",
    y = "Score")