Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Append Row to Rows (#4466) #4470

Merged
merged 4 commits into from
Jun 30, 2023
Merged

Append Row to Rows (#4466) #4470

merged 4 commits into from
Jun 30, 2023

Conversation

tustvold
Copy link
Contributor

Which issue does this PR close?

Closes #4466

Rationale for this change

What changes are included in this PR?

Are there any user-facing changes?

@github-actions github-actions bot added the arrow Changes to the arrow crate label Jun 30, 2023
@@ -832,14 +871,25 @@ struct RowConfig {
#[derive(Debug)]
pub struct Rows {
/// Underlying row bytes
buffer: Box<[u8]>,
buffer: Vec<u8>,
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I need to double-check this doesn't result in a performance regression, it shouldn't but stranger things have happened

Copy link
Contributor

@alamb alamb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you ❤️

Arc::ptr_eq(&row.config.fields, &self.config.fields),
"row was not produced by this RowConverter"
);
self.config.validate_utf8 |= row.config.validate_utf8;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why doesn't this just assert that the values of validate_utf8 are the same?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is consistent with the logic elsewhere, and is ultimately harmless

/// # use arrow_row::{Row, RowConverter, SortField};
/// # use arrow_schema::DataType;
/// #
/// let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]).unwrap();
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
/// let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]).unwrap();
/// // This example shows how to buffer only the Row values
/// let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]).unwrap();

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is already stated just above the code block, the commented out imports just make it hard to see

image

@tustvold
Copy link
Contributor Author

Unfortunately this does appear to regress performance, in some cases quite dramatically...

@tustvold
Copy link
Contributor Author

There is a fair bit of noise, but I'm happy this doesn't drastically regress the performance anymore

convert_columns 4096 u64(0)
                        time:   [17.309 µs 17.317 µs 17.328 µs]
                        change: [+0.3653% +1.2777% +2.3487%] (p = 0.02 < 0.05)
                        Change within noise threshold.
Found 17 outliers among 100 measurements (17.00%)
  1 (1.00%) low mild
  1 (1.00%) high mild
  15 (15.00%) high severe

convert_columns_prepared 4096 u64(0)
                        time:   [17.161 µs 17.166 µs 17.173 µs]
                        change: [-15.951% -15.847% -15.746%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
  3 (3.00%) high mild
  5 (5.00%) high severe

convert_rows 4096 u64(0)
                        time:   [43.738 µs 43.749 µs 43.762 µs]
                        change: [-9.0306% -8.7858% -8.6404%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
  1 (1.00%) low mild
  5 (5.00%) high mild
  3 (3.00%) high severe

convert_columns 4096 i64(0)
                        time:   [17.304 µs 17.311 µs 17.318 µs]
                        change: [-1.6188% -1.4897% -1.3000%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 15 outliers among 100 measurements (15.00%)
  4 (4.00%) low mild
  3 (3.00%) high mild
  8 (8.00%) high severe

convert_columns_prepared 4096 i64(0)
                        time:   [19.085 µs 19.316 µs 19.514 µs]
                        change: [+8.4826% +9.4567% +10.488%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 17 outliers among 100 measurements (17.00%)
  17 (17.00%) low severe

convert_rows 4096 i64(0)
                        time:   [45.207 µs 45.222 µs 45.242 µs]
                        change: [-6.3163% -6.1969% -5.9861%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe

convert_columns 4096 string(10, 0)
                        time:   [63.672 µs 63.687 µs 63.706 µs]
                        change: [+0.4749% +0.7652% +1.0325%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 12 outliers among 100 measurements (12.00%)
  7 (7.00%) high mild
  5 (5.00%) high severe

convert_columns_prepared 4096 string(10, 0)
                        time:   [63.474 µs 63.490 µs 63.505 µs]
                        change: [+0.4334% +0.6488% +0.7677%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 3 outliers among 100 measurements (3.00%)
  3 (3.00%) high mild

convert_rows 4096 string(10, 0)
                        time:   [62.774 µs 62.790 µs 62.810 µs]
                        change: [-3.8972% -3.8456% -3.7968%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
  1 (1.00%) low mild
  2 (2.00%) high mild
  7 (7.00%) high severe

convert_columns 4096 string(30, 0)
                        time:   [63.683 µs 63.701 µs 63.719 µs]
                        change: [+0.8005% +1.1281% +1.5777%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
  5 (5.00%) high mild
  2 (2.00%) high severe

convert_columns_prepared 4096 string(30, 0)
                        time:   [63.637 µs 63.662 µs 63.688 µs]
                        change: [+0.5443% +0.8511% +1.1496%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 1 outliers among 100 measurements (1.00%)
  1 (1.00%) high severe

convert_rows 4096 string(30, 0)
                        time:   [63.752 µs 64.440 µs 65.319 µs]
                        change: [+3.5137% +5.0297% +6.3617%] (p = 0.00 < 0.05)
                        Performance has regressed.

convert_columns 4096 string(100, 0)
                        time:   [85.907 µs 85.961 µs 86.021 µs]
                        change: [+4.2815% +4.6011% +4.9283%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 1 outliers among 100 measurements (1.00%)
  1 (1.00%) high severe

convert_columns_prepared 4096 string(100, 0)
                        time:   [85.606 µs 85.624 µs 85.645 µs]
                        change: [+3.9518% +4.4749% +4.8798%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) high mild
  5 (5.00%) high severe

convert_rows 4096 string(100, 0)
                        time:   [95.406 µs 95.517 µs 95.639 µs]
                        change: [+0.3985% +0.6151% +0.8626%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 6 outliers among 100 measurements (6.00%)
  3 (3.00%) low severe
  3 (3.00%) high severe

convert_columns 4096 string(100, 0.5)
                        time:   [102.21 µs 102.26 µs 102.33 µs]
                        change: [+4.1579% +4.3407% +4.5809%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 7 outliers among 100 measurements (7.00%)
  5 (5.00%) high mild
  2 (2.00%) high severe

convert_columns_prepared 4096 string(100, 0.5)
                        time:   [102.05 µs 102.10 µs 102.15 µs]
                        change: [+4.2080% +4.3656% +4.6188%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 8 outliers among 100 measurements (8.00%)
  1 (1.00%) low mild
  2 (2.00%) high mild
  5 (5.00%) high severe

convert_rows 4096 string(100, 0.5)
                        time:   [98.614 µs 98.660 µs 98.713 µs]
                        change: [-2.6106% -2.2896% -1.9902%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
  1 (1.00%) low mild
  2 (2.00%) high mild
  3 (3.00%) high severe

convert_columns 4096 string_dictionary(10, 0)
                        time:   [903.72 µs 903.96 µs 904.27 µs]
                        change: [+1.6355% +1.8429% +2.0210%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 6 outliers among 100 measurements (6.00%)
  3 (3.00%) high mild
  3 (3.00%) high severe

convert_columns_prepared 4096 string_dictionary(10, 0)
                        time:   [165.34 µs 165.39 µs 165.47 µs]
                        change: [+0.0240% +0.2124% +0.3262%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
  1 (1.00%) low severe
  2 (2.00%) high mild
  4 (4.00%) high severe

convert_rows 4096 string_dictionary(10, 0)
                        time:   [267.33 µs 267.39 µs 267.49 µs]
                        change: [-1.6229% -1.5810% -1.5314%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 11 outliers among 100 measurements (11.00%)
  2 (2.00%) low mild
  4 (4.00%) high mild
  5 (5.00%) high severe

convert_columns 4096 string_dictionary(30, 0)
                        time:   [918.56 µs 918.83 µs 919.11 µs]
                        change: [+0.5134% +0.6866% +0.8530%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 12 outliers among 100 measurements (12.00%)
  1 (1.00%) low mild
  7 (7.00%) high mild
  4 (4.00%) high severe

convert_columns_prepared 4096 string_dictionary(30, 0)
                        time:   [169.24 µs 169.28 µs 169.33 µs]
                        change: [-0.5830% -0.2718% -0.0845%] (p = 0.02 < 0.05)
                        Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
  2 (2.00%) low mild
  6 (6.00%) high mild
  5 (5.00%) high severe

convert_rows 4096 string_dictionary(30, 0)
                        time:   [269.76 µs 269.81 µs 269.88 µs]
                        change: [-2.4372% -2.2900% -2.0986%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 16 outliers among 100 measurements (16.00%)
  5 (5.00%) low mild
  5 (5.00%) high mild
  6 (6.00%) high severe

convert_columns 4096 string_dictionary(100, 0)
                        time:   [884.64 µs 884.82 µs 885.04 µs]
                        change: [+0.1187% +0.2086% +0.2745%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 13 outliers among 100 measurements (13.00%)
  2 (2.00%) low mild
  5 (5.00%) high mild
  6 (6.00%) high severe

convert_columns_prepared 4096 string_dictionary(100, 0)
                        time:   [221.10 µs 221.15 µs 221.22 µs]
                        change: [-0.1477% -0.0124% +0.1862%] (p = 0.91 > 0.05)
                        No change in performance detected.
Found 12 outliers among 100 measurements (12.00%)
  1 (1.00%) low mild
  7 (7.00%) high mild
  4 (4.00%) high severe

convert_rows 4096 string_dictionary(100, 0)
                        time:   [303.68 µs 303.79 µs 303.90 µs]
                        change: [-0.5015% -0.2210% +0.0415%] (p = 0.07 > 0.05)
                        No change in performance detected.
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe

convert_columns 4096 string_dictionary_non_preserving(100, 0)
                        time:   [165.64 µs 165.69 µs 165.74 µs]
                        change: [+3.0833% +3.3515% +3.6023%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 3 outliers among 100 measurements (3.00%)
  2 (2.00%) high mild
  1 (1.00%) high severe

convert_columns_prepared 4096 string_dictionary_non_preserving(100, 0)
                        time:   [162.89 µs 162.94 µs 162.99 µs]
                        change: [+2.7897% +3.0292% +3.2802%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 7 outliers among 100 measurements (7.00%)
  1 (1.00%) low severe
  3 (3.00%) high mild
  3 (3.00%) high severe

convert_rows 4096 string_dictionary_non_preserving(100, 0)
                        time:   [94.879 µs 94.969 µs 95.058 µs]
                        change: [-0.8412% -0.6973% -0.4752%] (p = 0.00 < 0.05)
                        Change within noise threshold.
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) low mild
  2 (2.00%) high mild
  3 (3.00%) high severe

convert_columns 4096 string_dictionary(100, 0.5)
                        time:   [423.63 µs 424.32 µs 425.71 µs]
                        change: [-7.9284% -6.7005% -5.4051%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe

convert_columns_prepared 4096 string_dictionary(100, 0.5)
                        time:   [142.48 µs 142.55 µs 142.62 µs]
                        change: [+2.3471% +2.6818% +3.0239%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 11 outliers among 100 measurements (11.00%)
  4 (4.00%) high mild
  7 (7.00%) high severe

convert_rows 4096 string_dictionary(100, 0.5)
                        time:   [156.65 µs 156.70 µs 156.76 µs]
                        change: [-3.3369% -3.0405% -2.7526%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 6 outliers among 100 measurements (6.00%)
  4 (4.00%) high mild
  2 (2.00%) high severe

convert_columns 4096 string_dictionary_non_preserving(100, 0.5)
                        time:   [121.10 µs 121.17 µs 121.25 µs]
                        change: [-4.4204% -4.1771% -3.9828%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
  5 (5.00%) high mild
  3 (3.00%) high severe

convert_columns_prepared 4096 string_dictionary_non_preserving(100, 0.5)
                        time:   [119.18 µs 119.20 µs 119.23 µs]
                        change: [-4.9679% -4.9138% -4.8618%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 7 outliers among 100 measurements (7.00%)
  4 (4.00%) high mild
  3 (3.00%) high severe

convert_rows 4096 string_dictionary_non_preserving(100, 0.5)
                        time:   [98.298 µs 98.328 µs 98.359 µs]
                        change: [-2.6643% -2.3476% -2.0359%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 3 outliers among 100 measurements (3.00%)
  1 (1.00%) high mild
  2 (2.00%) high severe

convert_columns 4096 string(20, 0.5), string(30, 0), string(100, 0), i64(0)
                        time:   [254.95 µs 255.19 µs 255.46 µs]
                        change: [-10.552% -10.378% -10.145%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 5 outliers among 100 measurements (5.00%)
  3 (3.00%) high mild
  2 (2.00%) high severe

convert_columns_prepared 4096 string(20, 0.5), string(30, 0), string(100, 0), i64(0)
                        time:   [254.65 µs 255.05 µs 255.52 µs]
                        change: [-9.5911% -9.3850% -9.1608%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
  2 (2.00%) high mild

convert_rows 4096 string(20, 0.5), string(30, 0), string(100, 0), i64(0)
                        time:   [258.38 µs 258.45 µs 258.52 µs]
                        change: [+2.3731% +2.6820% +2.9723%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 6 outliers among 100 measurements (6.00%)
  3 (3.00%) high mild
  3 (3.00%) high severe

convert_columns 4096 4096 string_dictionary(20, 0.5), string_dictionary(30, 0), string_dictionary(10...
                        time:   [453.39 µs 454.62 µs 455.71 µs]
                        change: [+2.1490% +2.7530% +3.4015%] (p = 0.00 < 0.05)
                        Performance has regressed.
Found 3 outliers among 100 measurements (3.00%)
  3 (3.00%) low mild

convert_columns_prepared 4096 4096 string_dictionary(20, 0.5), string_dictionary(30, 0), string_dict...
                        time:   [416.87 µs 417.58 µs 418.29 µs]
                        change: [-4.1962% -3.7800% -3.3997%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
  2 (2.00%) high mild

convert_rows 4096 4096 string_dictionary(20, 0.5), string_dictionary(30, 0), string_dictionary(100, ...
                        time:   [256.80 µs 257.17 µs 257.66 µs]
                        change: [-0.5142% +0.8155% +1.7484%] (p = 0.19 > 0.05)
                        No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
  7 (7.00%) low mild
  2 (2.00%) high mild
  4 (4.00%) high severe

@tustvold tustvold merged commit d7fa775 into apache:master Jun 30, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
arrow Changes to the arrow crate
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Request: a way to copy a Row to Rows
2 participants