Error in library(mda) : there is no package called 'mda' in the r_bring_your_own example

Hi there,

When running the r_bring_your_own example notebook, the training job is always failing due to the following error I see in the CW logs :
``` Error in library(mda) : there is no package called 'mda'```

Inspect the r_bring_your_own example notebook and reproduce its training job, then check the CloudWatch logs for the missing mda package. Done means the example training job completes without the “there is no package called 'mda'” error.

Written by the indexing model from the issue text.