This pull request adds a new sampler, FixedSample. This is useful for instances where you can load the entire dataset into memory, and want to make sampling decisions based on the total size of the dataset.
I also made some updates to CsvReader so it's more resilient to strangely formatted CSVs.
This pull request adds a new sampler,
FixedSample
. This is useful for instances where you can load the entire dataset into memory, and want to make sampling decisions based on the total size of the dataset.I also made some updates to
CsvReader
so it's more resilient to strangely formatted CSVs.