An example of this issue is the scenario in which you run the resampler and the data was stored with your dataframe being returned. Let's say you lose the instance that has the resampled data dataframe stored... you would have to rerun the resampler to produce the dataframe again. The resampler will process the resampling implementation again. This is inefficient. The resampler should be able to provide the dataframe without having to run through all of the data and resampling again.
An example of this issue is the scenario in which you run the resampler and the data was stored with your dataframe being returned. Let's say you lose the instance that has the resampled data dataframe stored... you would have to rerun the resampler to produce the dataframe again. The resampler will process the resampling implementation again. This is inefficient. The resampler should be able to provide the dataframe without having to run through all of the data and resampling again.