1. R setup

# Import libraries
library(tidyverse) # collection of R packages designed for data science
library(tidyr)
library(plotly)

# Number formatting
options(scipen = 1000000)
options(digits = 6)

2. import data


urlfile1="https://storage.googleapis.com/kantonzh-covid-hkfsaqgshw/gmaps_scrape/all_valid.csv"
dat1<-read_csv(url(urlfile1))

urlfile2 = "https://docs.google.com/spreadsheets/d/1KDqquW2axaUM9Z62JbyppuPq09IpAZRSIpPLb08nVqQ/pub?output=csv"
dat2 <-read_csv(urlfile2) %>% 
  filter(description != "Flughafen Zürich")

3. prepare data


dat_join <- dat1 %>%
  left_join(dat2, by = c("url" = "url")) 

# check: should be equal
## nrow(dat_join) == nrow(dat1)

4. plot data: 20.03.2020


dat <- dat_join %>% 
  filter(day_of_week == "Friday") %>% 
  select(description, 
         url, 
         hour_of_day, 
        'typical popularity over the last several weeks' = popularity_percent_normal, 
         '20.03.2020' = popularity_percent_current) %>%
  pivot_longer(
    cols = c('20.03.2020', 'typical popularity over the last several weeks'),
    names_to = "variable",
    values_to = "values") %>%
  filter(!is.na(description)) %>%
  arrange(description, hour_of_day)

# create graphing function
function.graph <- function(df){
  
# create looping variable
description_sel <- unique(dat$description)

# create for loop to produce ggplot2 graphs 
for (i in 1:length(description_sel)) { 
  
dat_plot <-  dat %>%
  filter(description == description_sel[i])

url_sel <- dat2 %>% 
  filter(description == description_sel[i]) %>%
  select(url)

#print(i)
#print(paste0("3333333333333333"))
#print(description_sel[i])
#print(as.character(url_sel))
#print(paste0("<a href='", url_sel, "'>", "link", "</a>"))

p <-ggplot(dat_plot, aes(hour_of_day, values)) +
  geom_col(stat = "identity", aes(fill = variable), position = "dodge") +
  scale_x_continuous(breaks= dat_plot$hour_of_day) +
  labs(title=paste0(description_sel[i]),
  #labs(title=paste0(description_sel[i], "\n", url_sel) ,
       subtitle = "Popularity for given hour is shown relative to typical peak popularity for location for week",
        x ="time",
        y = "")+
  theme(
    plot.title = element_text(size=28, colour = "#0033cc"), 
    axis.title.y=element_blank(),
    axis.text.y=element_blank(),
    axis.ticks.y=element_blank(),
    panel.border = element_blank(),
    panel.grid.major = element_blank(),
    panel.grid.minor = element_blank(),
    panel.background = element_blank(),
    legend.title=element_blank(),
    legend.position='top',
    legend.justification='left',
    legend.direction="vertical"
  )+
  scale_fill_manual(values=c("#E6B0AA", "#CCD1D1"))

# print plots to screen
print(p)

}} 

# run function.graph
function.graph(dat)

4. plot data: 21.03.2020


dat <- dat_join %>% 
  filter(day_of_week == "Saturday") %>% 
  select(description, 
         url, 
         hour_of_day, 
        'typical popularity over the last several weeks' = popularity_percent_normal, 
         '21.03.2020' = popularity_percent_current) %>%
  pivot_longer(
    cols = c('21.03.2020', 'typical popularity over the last several weeks'),
    names_to = "variable",
    values_to = "values") %>%
  filter(!is.na(description)) %>%
  arrange(description, hour_of_day)

# create graphing function
function.graph <- function(df){
  
# create looping variable
description_sel <- unique(dat$description)

# create for loop to produce ggplot2 graphs 
for (i in 1:length(description_sel)) { 
  
dat_plot <-  dat %>%
  filter(description == description_sel[i])

url_sel <- dat2 %>% 
  filter(description == description_sel[i]) %>%
  select(url)

#print(i)
#print(paste0("3333333333333333"))
#print(description_sel[i])
#print(as.character(url_sel))
#print(paste0("<a href='", url_sel, "'>", "link", "</a>"))

p <-ggplot(dat_plot, aes(hour_of_day, values)) +
  geom_col(stat = "identity", aes(fill = variable), position = "dodge") +
  scale_x_continuous(breaks= dat_plot$hour_of_day) +
  labs(title=paste0(description_sel[i]),
  #labs(title=paste0(description_sel[i], "\n", url_sel) ,
       subtitle = "Popularity for given hour is shown relative to typical peak popularity for location for week",
        x ="time",
        y = "")+
  theme(
    plot.title = element_text(size=28, colour = "#0033cc"), 
    axis.title.y=element_blank(),
    axis.text.y=element_blank(),
    axis.ticks.y=element_blank(),
    panel.border = element_blank(),
    panel.grid.major = element_blank(),
    panel.grid.minor = element_blank(),
    panel.background = element_blank(),
    legend.title=element_blank(),
    legend.position='top',
    legend.justification='left',
    legend.direction="vertical"
  )+
  scale_fill_manual(values=c("#E6B0AA", "#CCD1D1"))

# print plots to screen
print(p)

}} 

# run function.graph
function.graph(dat)