Closed daattali closed 7 years ago
It's not related to #99 (that's about loading data stored on S3 as, say, a CSV directly into a data frame).
This seems doable. There are reasons to preserve the existing structure (the objects in the list have S3 [in the R sense] classes, which makes it possible to pass them around as the value of object
arguments in other functions). How about if I add another function, get_bucket_df()
that returns a data frame instead?
Either that, or provide an argument to return as data.frame. I don't have a very strong opinion on which one since it's your package On Jan 14, 2017 2:29 PM, "Thomas J. Leeper" notifications@github.com wrote:
It's not related to #99 https://github.com/cloudyr/aws.s3/issues/99 (that's about loading data stored on S3 as, say, a CSV directly into a data frame).
This seems doable. There are reasons to preserve the existing structure (the objects in the list have S3 [in the R sense] classes, which makes it possible to pass them around as the value of object arguments in other functions). How about if I add another function, get_bucket_df() that returns a data frame instead?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/cloudyr/aws.s3/issues/102#issuecomment-272647081, or mute the thread https://github.com/notifications/unsubscribe-auth/AA6IFLfF2lKbh3Lxm3RavXntkmj5idbnks5rSSIngaJpZM4LjtnB .
This is potentially related/identical to #99 (I couldn't figure out if you're talking about this issue or not)
I previously used the RAmazonS3 package, and their
listBucket()
function returned a nice data frame, where each row had the info for one file in the bucket. It was really convenient. I want to upgrade to usingaws.s3
, and it's a bit annoying to receive the data as a list. From my (limited) experience,get_bucket()
always returns a list where each element has an identical structure. Was there a reason why a list was preferred over a dataframe if the data is always structured nicely?