The mgcv package includes the function bam that fits generalized additive models over larger datasets by reducing the required memory for the fitting procedure. Run a simulation on a dataset with 15...

The mgcv package includes the function bam that fits generalized additive models over larger datasets by reducing the required memory for the fitting procedure. Run a simulation on a dataset with 15 columns and 150, 000 rows. Compare the speed and memory footprint of mgcv::gam and mgcv::bam. Is there a noticeable difference in either?