Finally, we get to turn our attention to the other staple of the tidyverse, tidyr
.
Though this package offers more functionality, the main purpose of this package is to reshape data (convert from long to wide format) in a tidy manner.
Let’s recreate long
, a long format that contains the play counts for each year/month, using the following code:
> long <- tracks %>% + group_by(theyear=year(thedate), themonth) %>% + summarise(N=n()) > long # A tibble: 107 x 3 # Groups: theyear [?] theyear themonth N <dbl> <ord> <int> 1 2008 Jan 877 2 2008 Feb 984 3 2008 Mar 1486 4 2008 Apr 1101 ... # ... with 97 more rows
Now let’s get this into wide format with the different month in its own columns.
The tidyr equivalent of the dcast
function is spread
. As its arguments, it takes the data to transform, the column that contains the categories to be spread across different columns, and the value...