We were recently building a Shiny App in which we had to load data from a very large dataframe. It was directly impacting the app initialization time, so we had to look into different ways of reading data from files to R (in this case the customer provided csv files) and identify the best one.
The goal of my post is to compare:
utils, which was the standard way of reading csv files to R in RStudio,
readrwhich replaced the former method as a standard way of doing it in RStudio,
First, let’s generate some random data
and save the files on a disk to evaluate the load time. Besides the
csv format we will also need
Next, let’s check our files’ sizes:
As we can see both
feather format files take up much more storage space.
Csv takes up 6 times and
feather 4 more comparing to
Looking to learn more about importing data into R, this DataCamp tutorial covers all you need to know about importing simple text files to more advanced SPSS and SAS files.
We will use
microbenchmark library to compare the read times of the following methods:
in 10 rounds.
And the winner is…
feather! However, using
feather requires prior conversion of the file to the feather format.
readRDS can improve performance (second and third place in terms of speed) and has an added benefit of storing smaller/compressed file. In both cases you will first have to convert your file to the proper format.
When it comes to reading from
fread significantly beats
read.csv, and thus is the best option to read a
We decided to go with
feather file since converting from
csv to this format is just a one time job and we didn’t have a strict limitation on a storage space to consider usaging of
The final workflow was:
csvfile provided by our customer using
featherfile on app initialization using
The first two tasks were done once and outside of a Shiny App context.
There is also quite an interesting benchmark done by Hadley here on reading complete files to R. Unfortunately, if you use functions defined in that post, you will end up with a character type object and will have to apply string manipulations to obtain a commonly and widely used dataframe.