readData.R

# Quick script to read in the file 
library(tidyverse)
library(RCurl) # still trying to get this one to work
library(bulkreadr) # needs to link to gmail account... which might make some people uncomfortable

# URL for the sheet
thisURL <- "https://docs.google.com/spreadsheets/d/1tlkZVdZP41NGNjXFQW84sBhe4AK0OMqjDwrS2FUgfbE/edit?gid=0#gid=0"

# doesn't work just now... probably because we're not pointing it to a .csv file
# download.file(thisURL, "temp.csv", method = 'curl')

# this works, needs to be linked to gmail... which isn't the worst thing ever
df_pokemon <- googlesheets4::read_sheet(thisURL, 
                           sheet = "Packs - Overview")

# tidy up the data... this person has clearly never worked on a data science project 
# but that's ok
theseColnames <- df_pokemon[4,2:ncol(df_pokemon)] %>% 
  gather(1:ncol(.), 
        key = "dontneed", 
        value = "names") %>%
  pull(names) %>%
  unique() %>% 
  unlist()

# NOTE: we should be careful about this part as it involves some hardcoding... and that could change things... 
# more tidying... 
theseColnames <- str_remove_all(theseColnames, " ")
theseColnames <- paste(rep(theseColnames, each = 2), c("HEAVY", "LIGHT"), sep = "_")

# get only the rows and columns we care about
df_cleaner <- df_pokemon[6:nrow(df_pokemon), 2:ncol(df_pokemon)]
colnames(df_cleaner) <- theseColnames

# make the data look nice
df_cleaner <- df_cleaner %>% 
  gather(everything(),
         key = "temp",
         value = "weight") %>%
  unnest(weight) %>% 
  separate(temp, 
           into = c("packType", "weightType"), 
           sep = "_")

df_cleaner %>% 
  ggplot(aes(weight, fill = weightType)) + 
  geom_histogram(aes(y = after_stat(density)), 
                 position = "dodge") + 
  facet_wrap(~packType, 
             scales = "free") + 
  see::scale_fill_flat()

ggsave("scratch/plots/firstPlot.png", 
       height = 3000, 
       width = 6000, 
       units = "px")