Respect copycols when building DataFrame from Tables.CopiedColumns #2656
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Fixes apache/arrow-julia#146. The issue here was we were just ignoring `copycols` entirely when building DataFrame from `Tables.CopiedColumns`. Which is mostly fine because `CopiedColumns` implies DataFrames can assume ownership of the columns. The wrinkle comes in when you consider cases like `Arrow.Table`, which wraps columns in `Tables.CopiedColumns`, not necessarily because copies have been made, but because the data is immutable, so it should be safe for other tables to assume ownership. The problem is when users want to _mutate_ those columns, they need to make copies of them, so naturally users try `DataFrame(::Arrow.Table; copycols=true)`, but then that doesn't actually make copies. The proposal here just intercepts the top-level `DataFrame(::CopiedColumns)` constructor and passes `copycols=false` as the default, so if users pass `copycols=true`, it will still be respected.
bkamins
reviewed
Mar 12, 2021
bkamins
approved these changes
Mar 12, 2021
bkamins
added a commit
that referenced
this pull request
Mar 31, 2021
…2656) * Respect copycols when building DataFrame from Tables.CopiedColumns Fixes apache/arrow-julia#146. The issue here was we were just ignoring `copycols` entirely when building DataFrame from `Tables.CopiedColumns`. Which is mostly fine because `CopiedColumns` implies DataFrames can assume ownership of the columns. The wrinkle comes in when you consider cases like `Arrow.Table`, which wraps columns in `Tables.CopiedColumns`, not necessarily because copies have been made, but because the data is immutable, so it should be safe for other tables to assume ownership. The problem is when users want to _mutate_ those columns, they need to make copies of them, so naturally users try `DataFrame(::Arrow.Table; copycols=true)`, but then that doesn't actually make copies. The proposal here just intercepts the top-level `DataFrame(::CopiedColumns)` constructor and passes `copycols=false` as the default, so if users pass `copycols=true`, it will still be respected. * Update src/other/tables.jl Co-authored-by: Bogumił Kamiński <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes apache/arrow-julia#146. The issue here
was we were just ignoring
copycolsentirely when building DataFramefrom
Tables.CopiedColumns. Which is mostly fine becauseCopiedColumnsimplies DataFrames can assume ownership of the columns.The wrinkle comes in when you consider cases like
Arrow.Table, whichwraps columns in
Tables.CopiedColumns, not necessarily because copieshave been made, but because the data is immutable, so it should be safe
for other tables to assume ownership. The problem is when users want to
mutate those columns, they need to make copies of them, so naturally
users try
DataFrame(::Arrow.Table; copycols=true), but then thatdoesn't actually make copies.
The proposal here just intercepts the top-level
DataFrame(::CopiedColumns)constructor and passescopycols=falseasthe default, so if users pass
copycols=true, it will still berespected.