01_saving data_to_disk.R

# saving a dataset in different formats
library(tidyverse)
library(palmerpenguins)
library(waldo)

df <- penguins_raw

# saving as csv
write_csv(df, "processed_data/penguins.csv")
file.size("processed_data/penguins.csv") / 1000

# serializing to compressed .rds
write_rds(df, "processed_data/penguins.rds", compress = "xz")
file.size("processed_data/penguins.rds") / 1000

# reading from .rds ensures exact replication
df_fromdisk <- read_rds("processed_data/penguins.rds")
identical(df, df_fromdisk)

# reading from csv means stuff has to be inferred - losing info
df_fromcsv <- read.csv("processed_data/penguins.csv")
identical(df, df_fromcsv)
compare(df, df_fromcsv)