This PR will make it so we can plot multiple years for #79
In order to do this it required a refactor of the code that generated the lag covariates, as the tidyr approach pulled too much RAM and made everything fall over. @goldingn provided a great approach that is by my estimates, at least 7 times less memory intensive at least 5 times faster for our data size.
Still working on developing the code to work for the pixel maps, but this should work soon.
This PR will make it so we can plot multiple years for #79
In order to do this it required a refactor of the code that generated the lag covariates, as the tidyr approach pulled too much RAM and made everything fall over. @goldingn provided a great approach that is by my estimates, at least 7 times less memory intensive at least 5 times faster for our data size.
Still working on developing the code to work for the pixel maps, but this should work soon.