Antoine Pitrou comments

Results 823 comments of


                                            Antoine Pitrou

GH-38007: [C++] Add VariableShapeTensor implementation

@rok I think this would be easier to review if you first submitted a purely C++ PR, can you do that?

[R][C++] Repartitioning on a new variable uses all my RAM and crashes

Issue resolved by pull request 40722 https://github.com/apache/arrow/pull/40722

GH-20213: [C++] Implement cast to/from halffloat

Thanks for posting this. The approach is sane, I posted some comments. You will have to decide whether you want to tackle casting to string, or prefer leaving that for...

GH-20213: [C++] Implement cast to/from halffloat

> @pitrou you mentioned string cast(s). > Is that: > 1. float16 to string > 2. string to float16 > ? Yes. > What other casts are required to consider...

GH-39377: [C++] IO: Reuse same buffer in CompressedInputStream

You should benchmark using a faster codec such as LZ4, if you really want to measure the overhead of CompressedInputStream.

GH-20213: [C++] Implement cast to/from halffloat

> There doesn't seem to be a direct way to cast from a half float ultimately represented by a `uint16_t` to a string, but I wonder if it's acceptable to...

GH-20213: [C++] Implement cast to/from halffloat

By the way, you'll have to implement unit tests at some point (in C++).

GH-39377: [C++] IO: Reuse same buffer in CompressedInputStream

It's ok @mapleFU !

GH-20213: [C++] Implement cast to/from halffloat

> So one question I currently have is, should I specialize `Datum` on `HalfFloatType` (or similar)? Which method would you specialize?

GH-20213: [C++] Implement cast to/from halffloat

> After reading more code, more likely I need to specialize or modify `CompareFloating`, `VisitFloatingEquality`, etc. But I'm not certain. If you're struggling on this, someone else can take a...