figure_2.Rmd

---
title: "figure_2"
output: html_document
---
Setup 
```{r}
library(scales)
library(knitr)
library(dplyr)
library(tidyverse)
library(ggplot2)

```



Loading mobility data
```{r}
load('/data/safegraph/safegraph.Rdata')
#View(safegraph)
nyc_zip_codes <- read_csv("https://raw.githubusercontent.com/erikgregorywebb/nyc-housing/master/Data/nyc-zip-codes.csv")
```

Getting pre pandemic and postpandemic data 
```{r}
safegraph <- safegraph[safegraph$postal_code %in% nyc_zip_codes$ZipCode, ]

pre_pand <- safegraph %>% filter(grepl("2020-02-", date)) 
View(pre_pand)

post_pand <- safegraph %>% filter(grepl("2020-03-", date) | grepl("2020-04-", date))
View(post_pand)
```

Create baselines 
```{r}
# BASELINE: What is the typical traffic in each zip --> meadian of average visits per day across all days 
baselines <- pre_pand %>% group_by(postal_code) %>% summarize(baseline = median(avg_visits_per_day)) %>% filter(baseline != 0)
View(baselines)
```

```{r}
df <- inner_join(post_pand, baselines, "postal_code") 
df <- df %>% mutate(delta = (avg_visits_per_day - baseline)/baseline) %>% filter(delta>=-1 & delta<=2)
df2 <- df %>% select(date, delta) %>% arrange(date) %>% group_by(date)%>% 
  filter(delta == median(delta))
  
View(df2)

```

Build the graph
```{r}
ggplot(data = df, mapping = aes(x = as.factor(date), y = delta)) +
  geom_violin()+
  geom_boxplot(data = df2, fill="red") +
  #geom_point(data = df2)+
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))+
  xlab("Date")+
  ylab("Change in mobility relative to baseline")
```
```{r}
save(data = baselines, post_pand, df, df2, file = "mobility.Rdata")
```