Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485:
URL: https://github.com/apache/arrow/issues/46485#issuecomment-2907728585
> Can we use `Fn` or something instead? (`*data.statistics` doesn't work
when `data.statistics.get() == nullptr`.)
>
> diff --git a/cpp/src/arrow/array/data.cc b/cpp/src/arrow/array/data.cc
> index 2e55668fb9..6da4bb5317 100644
> --- a/cpp/src/arrow/array/data.cc
> +++ b/cpp/src/arrow/array/data.cc
> @@ -165,7 +165,11 @@ Result> CopyToImpl(const
ArrayData& data,
> ARROW_ASSIGN_OR_RAISE(output->dictionary,
CopyToImpl(*data.dictionary, to, copy_fn));
>}
>
> - output->statistics = data.statistics;
> + if (std::is_same_v) {
> +output->statistics = data.statistics;
> + } else {
> +output->statistics =
std::make_shared(*data.statistics);
> + }
>
>return output;
> }
Because of this, the following line is not completely correct. However,
let's discuss it in another issue that I will create soon
https://github.com/apache/arrow/blob/153da3053b62f8f9bcafb87546e12a979e605734/cpp/src/arrow/array/data.cc#L179-L182
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
kou commented on issue #46485:
URL: https://github.com/apache/arrow/issues/46485#issuecomment-2907722275
Can we use `Fn` or something instead?
(`*data.statistics` doesn't work when `data.statistics.get() == nullptr`.)
```diff
diff --git a/cpp/src/arrow/array/data.cc b/cpp/src/arrow/array/data.cc
index 2e55668fb9..6da4bb5317 100644
--- a/cpp/src/arrow/array/data.cc
+++ b/cpp/src/arrow/array/data.cc
@@ -165,7 +165,11 @@ Result> CopyToImpl(const
ArrayData& data,
ARROW_ASSIGN_OR_RAISE(output->dictionary, CopyToImpl(*data.dictionary,
to, copy_fn));
}
- output->statistics = data.statistics;
+ if (std::is_same_v) {
+output->statistics = data.statistics;
+ } else {
+output->statistics =
std::make_shared(*data.statistics);
+ }
return output;
}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485:
URL: https://github.com/apache/arrow/issues/46485#issuecomment-2907631297
> Given this logic, it seems the current implementation should be modified
to copy the arrow::ArrayStatistics instead of creating a new one during the
copy operation.
1-Sorry for the confusion in my above message. What I meant to say is that
the current implementation should be updated to copy the
`arrow::ArrayStatistics `rather than share it during the copy operation.
Note: `arrow::ArrayData::statistics` is a `std::shared_ptr `, so the
following line shares the object instead of copying it.
https://github.com/apache/arrow/blob/8afaf95b38459f97ab12b22ebb7e1913b31025bf/cpp/src/arrow/array/data.cc#L167-L169
2- Both `arrow::ArrayData::ViewOrCopy` and `arrow::ArrayData::CopyTo `use
[CopyToImpl](https://github.com/apache/arrow/blob/153da3053b62f8f9bcafb87546e12a979e605734/cpp/src/arrow/array/data.cc#L146-L171).
Therefore, it should be specified whether the array is being copied in the
line below. If the array is not copied, then `arrow::ArrayData::statistics
`should be shared (which is the current behavior). Otherwise, it should be
copied.
https://github.com/apache/arrow/blob/8afaf95b38459f97ab12b22ebb7e1913b31025bf/cpp/src/arrow/array/data.cc#L167-L169
My suggestion is to use the following code instead of the one above
```c++
if (output->buffers[0]->address() == data.buffers[0]->address()) {
// The output is a view
output->statistics = data.statistics;
}else {
// The output is a copy
output->statistics = std::make_shared(*data.statistics);
}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
kou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2907563853 1. What is the difference of "copy" and "creating a new one" in " copy the` arrow::ArrayStatistics` instead of creating a new one during the copy operation"? I think "copy" create a new one with the same content. 2. Where do we need it? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2903984104 @Kou, I have several questions regarding the` CopyTo` and` ViewOrCopy` operations. I understand that `arrow::Array` is designed to be immutable, and based on that design, it's acceptable to share buffer ownership among multiple arrays (as I've seen in the casting compute function). However, when performing a deep copy on an `arrow::Array` , the intent is to produce a new array with its own unique buffer. Given this logic, it seems the current implementation should be modified to copy the` arrow::ArrayStatistics` instead of creating a new one during the copy operation. https://github.com/apache/arrow/blob/8afaf95b38459f97ab12b22ebb7e1913b31025bf/cpp/src/arrow/array/data.cc#L167-L169 1- Is my observation correct 2- Can the following code be used to distinguish between View and Copy operations on an `arrow::Buffer`? ```c++ buffer_in->data() == buffer_out->data() ``` By the way, if the answer to question one is yes, I'd like to address that first -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2897972267 take -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
pitrou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893878388 Yes, please open an issue, thank you! -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893796275 > > This happens because `fixed_size_list` uses a single bitmap layout. Consequently, the`int32`bitmap layout is ignored in the below code, and only one data layout remains — which coincidentally matches the layout of int32. > > Ah, you're right. That code is certainly too naive... @pitrou Should I open a separate issue for this? I may not be able to address it in the near future due to both my current knowledge limitations and the number of other issues I'm already involved with -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893786141 > > Ah, wait. Is statistics still valid with different data type...? > > Certainly not. If you view a `int32` array as `uint32`, then the min and max may be different if there are negative integers. Our focus is on preserving the `statistics attributes `as much as possible, as I mentioned [here](https://github.com/apache/arrow/issues/46485#issuecomment-2888790138)—for example, `ARROW:average_byte_width:exact`. However, with the current behavior of `arrow::Array::View`, it is not possible to preserve them. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
pitrou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893747015 > This happens because `fixed_size_list` uses a single bitmap layout. Consequently, the`int32`bitmap layout is ignored in the below code, and only one data layout remains — which coincidentally matches the layout of int32. Ah, you're right. That code is certainly too naive... -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893733972 > > Hmm. The important part is that `fixed_size_list(int32(), 2)` -view-> `list_view(int32())` -view-> `fixed_size_list(int32(), 2)` doesn't work, right? [@pitrou](https://github.com/pitrou) What do you think about this behavior? > > I don't know how it would be possible to view a fixed-size list as a variable-size list, even though there is no offsets or views buffer available. It should probably return a TypeError. @pitrou As I mentioned [here](https://github.com/apache/arrow/issues/46485#issuecomment-2893193246), no type error is emitted during the conversion from` list_view(fixed_size_list(int32, 2))` to `list_view(int32),` even when the array is built with only half the expected number of values (0..8). This happens because `fixed_size_list` uses a single bitmap layout. Consequently, the` int32 `bitmap layout is ignored in the code, and only one data layout remains — which coincidentally matches the layout of int32. https://github.com/apache/arrow/blob/a1707dba7c9ec4265c048586c485c526048cf162/cpp/src/arrow/array/data.cc#L873-L880 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893719092 > > Hmm. The important part is that `fixed_size_list(int32(), 2)` -view-> `list_view(int32())` -view-> `fixed_size_list(int32(), 2)` doesn't work, right? [@pitrou](https://github.com/pitrou) What do you think about this behavior? > > I don't know how it would be possible to view a fixed-size list as a variable-size list, even though there is no offsets or views buffer available. It should probably return a TypeError. @pitrou As I mentioned [here](https://github.com/apache/arrow/issues/46485#issuecomment-2893193246), no type error is emitted, even when the array is built with only half the expected number of values (0..8). This is because fixed_size_list uses a single bitmap layout. As a result, the int32() bitmap layout is ignored in the code below, and only one data layout remains — which happens to match the int32 layout. https://github.com/apache/arrow/blob/a1707dba7c9ec4265c048586c485c526048cf162/cpp/src/arrow/array/data.cc#L873-L880 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
pitrou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893640268 > Hmm. The important part is that `fixed_size_list(int32(), 2)` -view-> `list_view(int32())` -view-> `fixed_size_list(int32(), 2)` doesn't work, right? [@pitrou](https://github.com/pitrou) What do you think about this behavior? I don't know how it would be possible to view a fixed-size list as a variable-size list, even though there is no offsets or views buffer available. It should probably return a TypeError. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
pitrou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893637216 > Ah, wait. Is statistics still valid with different data type...? Certainly not. If you view a `int32` array as `uint32`, then the min and max may be different if there are negative integers. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2893193246 > Hmm. The important part is that `fixed_size_list(int32(), 2)` -view-> `list_view(int32())` -view-> `fixed_size_list(int32(), 2)` doesn't work, right? [@pitrou](https://github.com/pitrou) What do you think about this behavior? There are two issues: 1-During the conversion from `fixed_size_list(int32(), 2)` to `list_view(int32())`, half of the data is lost — only the first 9 elements (`0 to 8`) are visible in` list_view(int32())`, even though the original range was `0 to 17`. 2-As you mentioned, converting from `list_view(int32()) `back to `fixed_size_list(int32(), 2)` results in an invalid array. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
kou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2892605420 > Doesn't the following code risk causing a stack overflow? Shouldn't a depth limit be considered to prevent that? Because we use recursive function calls here, right? In general, a data type for view is provided by a user. It can be controlled by a user. If a user gets the stack overflow error, a user will report it to us. So I think that we don't need to care about it for now. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
kou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2892596991 > 2-Should I think of this method as a form of casting based on `arrow::DataTypeLayout`? Yes. > 3-Is the following output for `view` and `view_2` acceptable?(If yes, I think it might make sense to drop `arrow::ArrayStatistics`.) Hmm. The important part is that `fixed_size_list(int32(), 2)` -view-> `list_view(int32())` -view-> `fixed_size_list(int32(), 2)` doesn't work, right? @pitrou What do you think about this behavior? Note that `Array::View` was implemented by #4482. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2890156783 I have another question that's not directly related to this issue, but it's about arrow::Array::View. Doesn't the following code risk causing a stack overflow? Shouldn't a depth limit be considered to prevent that? https://github.com/apache/arrow/blob/7f645d404a16e8c7c939dec70ad61f4ae4de7730/cpp/src/arrow/array/data.cc#L739-L753 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485:
URL: https://github.com/apache/arrow/issues/46485#issuecomment-2890083612
@kou
I have two questions regarding `arrow::Array::View`:
2-Should I think of this method as a form of casting based on
`arrow::DataTypeLayout`?
3-Is the following output for `view` and `view_2` acceptable?(If yes, I
think it might make sense to drop `arrow::ArrayStatistics`.)
```c++
Result> MakeListView() {
auto list_view_type = list_view(fixed_size_list(int32(), 2));
auto builder = MakeBuilder(list_view_type).ValueOrDie();
auto list_view_builder =
internal::checked_pointer_cast(std::move(builder));
auto fixed_size_builder =
internal::checked_cast(list_view_builder->value_builder());
auto int32_builder =
internal::checked_cast(fixed_size_builder->value_builder());
for (int32_t i = 0; i < 9; ++i) {
if (i % 3 == 0) {
ARROW_RETURN_NOT_OK(list_view_builder->Append(true, 3));
}
ARROW_RETURN_NOT_OK(fixed_size_builder->Append());
}
ARROW_RETURN_NOT_OK(list_view_builder->AppendNull());
for (int i = 0; i < 18; i++) {
ARROW_RETURN_NOT_OK(int32_builder->Append(i));
}
return list_view_builder->Finish();
}
TEST(View, Test) {
ASSERT_OK_AND_ASSIGN(auto array, MakeListView());
auto view = array->View(list_view(int32())).ValueOrDie();
ARROW_LOGGER_INFO("", array->ToString());
ARROW_LOGGER_INFO("", view->ToString());
auto view_2 = view->View(list_view(fixed_size_list(int32(),
2))).ValueOrDie();
ARROW_LOGGER_INFO("", view_2->ToString());
}
```
output
```
// ARROW_LOGGER_INFO("", array->ToString())
[
[
0,
1
],
[
2,
3
],
[
4,
5
]
],
[
[
6,
7
],
[
8,
9
],
[
10,
11
]
],
[
[
12,
13
],
[
14,
15
],
[
16,
17
]
],
null
]
// ARROW_LOGGER_INFO("", view->ToString());
[
[
0,
1,
2
],
[
3,
4,
5
],
[
6,
7,
8
],
null
]
// ARROW_LOGGER_INFO("", view_2->ToString());
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2888790138 Statistical attributes might be grouped into two types: target-independent attributes, like` ARROW:average_byte_width:exact,` and target-dependent attributes, such as `min` and `max`. It seems possible to preserve target-independent attributes. I’ll look into it more and share my final thoughts -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
kou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2888741367 Ah, wait. Is statistics still valid with different data type...? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
kou commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2888740860 Good catch. We don't need to discard `arrow::ArrayStatistics` by `arrow::Array::View()`. Could you open a new issue for it? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [I] [C++][Statistics][Docs] Clarify whether arrow::ArrayStatistics is discarded during View, Slice, and Copy operations in the documentation for arrow::Array and arrow::ArrayData [arrow]
andishgar commented on issue #46485: URL: https://github.com/apache/arrow/issues/46485#issuecomment-2888472845 @kou, arrow::ArrayStatistics is discarded in arrow::Array::View. Is this the expected behavior? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
