COVID-19 vaccine interest in Malaysia

We are going to do a basic google trends search using gtrendsR package and do some plotting with ggplot2.

These are the required packages.

library(gtrendsR)
library(tidyverse)

Run gtrends() function to search our keywords of interest (i.e; type of vaccine). So far, we only used 4 type of vaccines in Malaysia.

vaccine <- gtrends(c("pfizer", "astrazeneca", "sinovac", "cansino"), geo = "MY")

Then, plot our keywords.

plot(vaccine)

Probably, it’s better if we filter our date to when the COVID-19 pandemic started, which is around March 2020.

vaccine$interest_over_time %>% 
  group_by(keyword) %>% 
  filter(hits != "<1" & date > as.Date("2020-03-01")) %>% 
  mutate(hits = as.numeric(hits), 
         date = as.Date(date)) %>% 
  ggplot() + 
  geom_line(aes(x = date, y = hits, color = keyword), size = 0.8) +
  theme_minimal() +
  labs(title = "COVID-19 vaccine interest in Malaysia", y = "Search hits", x = "Date") +
  scale_x_date(date_breaks = "4 month")

So, AstraZeneca vaccine is of high interest, probably due to infamous blood clotting issue. Next, we can also get the search keywords based on the states.

vaccine$interest_by_region %>% 
  group_by(location) %>% 
  ggplot(aes(location, hits, fill = keyword)) +
  geom_col(alpha = 0.8) +
  coord_flip() +
  theme_minimal() +
  scale_fill_viridis_d() +
  labs(title = "COVID-19 vaccine interest in Malaysia by states", y = "Search hits", x = "")

Lastly, we can plot the search keywords based on the city.

vaccine$interest_by_city %>% 
  group_by(location) %>% 
  drop_na() %>% 
  ggplot(aes(location, hits, fill = keyword)) +
  geom_col(alpha = 0.8) +
  coord_flip() +
  theme_minimal() +
  scale_fill_viridis_d() +
  labs(title = "COVID-19 vaccine interest in Malaysia by cities", y = "Search hits", x = "")

gtrendsR with just a bit of plots certainly very useful if we want to gauge certain issues in the community.