Confused about using data.table package,
Hi Carl, I have not fully learned dplyr, but it seems harder than tapply() and the ?apply() family in general. Almost every ggplot2 data I have seen is manipulated using dplyr. Something must be good about dplyr. aggregate(), tapply(), do.call(), rbind() will be sorely missed! :( Thanks!
On Tue, Feb 21, 2017 at 4:21 PM, Carl Sutton <suttoncarl at ymail.com> wrote:
Hi I have found that: A) Hadley's new book to be wonderful on how to use dplyr, ggplot2 and his other packages. Read this and using as a reference saves major frustration. b) Data Camps courses on ggplot2 are also wonderful. GGPLOT2 has more capability than I have mastered or needed. To be an expert with ggplot2 will take some effort. To just get run of the mill helpful, beautiful plots, no major time needed for that. I use both of these sources regularly, especially when what is in my grey matter memory banks is not working. Refreshers are sometimes needed. If your data sets are large and available memory limited, then data.table is the package I use. I am amazed at the difference of memory usage with data.table versus other packages. My laptop has 16gb ram, and tidyr maxed it but data.table melt used less than 6gb(if I remember correctly) on my current work. Since discovering fread and fwrite, read.table, read.csv, and write have been benched. Every script I have includes library(data.table) Carl Sutton