let the data.frame be data.frame, and data.table be data.table.

[GreenEric]

Hi, Xiaodaign, Thanks for the great work. I noticed that, due to the legacy issue from some other package, the data.frame is more pleasible for data.frame, than data.table. What if I save/read the data as data.fram, and convert it back to data.table within the r-script, will it be workable? Thanks for your advice.

[xiaodaigh]

I actually don't understand what you are asking here. Please elaborate

[xiaodaigh]

I think you want it so that disk.frame doesn't auto convert data.frames to data.table. Is that correct?

[iqis]

@GreenEric I think you could find work around this issue pretty easily, as a data.table is a data.frame

[xiaodaigh]

That is true. It's still nice to be able to preserve that in a future release.

[xiaodaigh]

Let the user choose their preferred data.frames which are "base"=plain "data.frame" "data.table" or "tibble", so the returned data.frame is in the format most preferred by the user.