Product Hunt products

library(tidyverse)
product_hunt <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-10-04/product_hunt.csv')

product_hunt %>% head()

product_hunt %>%
  mutate(month=zoo::as.yearmon(as.Date(release_date)))%>%  # "%Y-%m-%d"
  count(month) %>%
  ggplot(aes(month,n)) +
  geom_point() 

product_hunt %>%
  mutate(month=zoo::as.yearmon(as.Date(release_date))) %>%  # "%Y-%m-%d"
  count(month) %>%
  ggplot(aes(month,n)) +
  geom_point() +
  geom_smooth()

product_hunt%>%
   mutate(month=zoo::as.yearmon(as.Date(release_date)))%>%
   count(name, month,product_ranking,upvotes) %>%
   drop_na() %>%
   arrange(-product_ranking,-upvotes)%>%
   group_by(product_ranking) %>%
  count(name)
   filter(upvotes==max(upvotes))
  

#column_to_rownames(name)

df <- product_hunt%>%
   mutate(month=zoo::as.yearmon(as.Date(release_date)))%>%
   count(name, month,product_ranking,upvotes) %>%
   drop_na() %>%
   arrange(-product_ranking,-upvotes)%>%
   group_by(product_ranking) %>%
   filter(upvotes==max(upvotes))

df

product_hunt%>%
   #mutate(month=zoo::as.yearmon(as.Date(release_date)))%>%
  mutate(year=lubridate::year(as.Date(release_date)))%>%
   count(year,product_ranking,upvotes) %>%
   drop_na() %>%
   arrange(-product_ranking,-upvotes) %>%
  group_by(year,product_ranking)%>%
  summarise(avg_votes=sum(upvotes)) %>%
  ungroup() %>%
  mutate(year=factor(year)) 

product_hunt%>%
   #mutate(month=zoo::as.yearmon(as.Date(release_date)))%>%
  mutate(year=lubridate::year(as.Date(release_date)))%>%
   count(year,product_ranking,upvotes) %>%
   drop_na() %>%
   arrange(-product_ranking,-upvotes) %>%
  group_by(year,product_ranking)%>%
  summarise(avg_votes=sum(upvotes)) %>%
  ungroup() %>%
  mutate(year=factor(year)) %>%
  ggplot(aes(x = product_ranking, y = avg_votes,color=year,fill=year)) +
  geom_col() 

df <- product_hunt %>% 
  arrange(release_date) %>%
  select(release_date,upvotes,product_ranking) %>%
  group_by(release_date,product_ranking) %>%
  summarise(year = year(release_date),
            month = month(release_date, label=TRUE),
            day = day(release_date),
            tot_votes=scale(upvotes)) %>%
  ungroup() %>% 
  select(release_date,day,month,year,tot_votes,product_ranking)%>%
        fill(tot_votes,product_ranking) %>%
  distinct() %>%
  drop_na() 

df

ggplot(df,aes(month,product_ranking,fill=tot_votes))+
 # geom_hex() +
  geom_tile(color="grey40",size=0.5) +
  scale_fill_viridis(name="Votes",option ="E",begin = 0.1,end = 1,direction = -1)+
  facet_grid(vars(year)) +
  scale_y_continuous(trans = "reverse", breaks = unique(df$day)) +
  scale_x_discrete(expand = c(0,0)) +
  labs(x="Month",y="Product ranking",
       title="Heatmap of the products upvotes by ranking in time",
       subtitle="darker areas show a higher number of votes (scaled values)",
       caption="DataSource: #TidyTuesday week40 Product Hunt | DataViz: Federica Gazzelloni") +
  theme_minimal(base_size = 9,base_family = "Roboto Condensed")+
  theme(text=element_text(color="grey90",face="bold"),
        axis.line = element_blank(),
        strip.switch.pad.grid = unit(2,units = "pt"),
        plot.background = element_rect(color="grey60",fill="grey60"),
        plot.title = element_text(size=18))

ggsave("w40_product_hunt.png",
       dpi=600,
       width = 10,
       height = 7)