Skip to content

Match table statistics with table schema #571

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.

Sign up for GitHub

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

Merged
merged 2 commits into from
Dec 15, 2023
Merged

Match table statistics with table schema #571

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
70952d3
Fix statistics
trueleo Dec 14, 2023
edb4dd5
Refactor
trueleo Dec 14, 2023
File filter

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
36 changes: 36 additions & 0 deletions server/src/catalog/column.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,8 @@

use std::cmp::{max, min};

use arrow_schema::DataType;
use datafusion::scalar::ScalarValue;
use parquet::file::statistics::Statistics;

#[derive(Debug, Clone, serde::Serialize, serde::Deserialize)]
Expand Down Expand Up @@ -85,6 +87,40 @@ impl TypedStatistics {
_ => panic!("Cannot update wrong types"),
}
}

pub fn min_max_as_scalar(self, datatype: &DataType) -> Option<(ScalarValue, ScalarValue)> {
let (min, max) = match (self, datatype) {
(TypedStatistics::Bool(stats), DataType::Boolean) => (
ScalarValue::Boolean(Some(stats.min)),
ScalarValue::Boolean(Some(stats.max)),
),
(TypedStatistics::Int(stats), DataType::Int32) => (
ScalarValue::Int32(Some(stats.min as i32)),
ScalarValue::Int32(Some(stats.max as i32)),
),
(TypedStatistics::Int(stats), DataType::Int64) => (
ScalarValue::Int64(Some(stats.min)),
ScalarValue::Int64(Some(stats.max)),
),
(TypedStatistics::Float(stats), DataType::Float32) => (
ScalarValue::Float32(Some(stats.min as f32)),
ScalarValue::Float32(Some(stats.max as f32)),
),
(TypedStatistics::Float(stats), DataType::Float64) => (
ScalarValue::Float64(Some(stats.min)),
ScalarValue::Float64(Some(stats.max)),
),
(TypedStatistics::String(stats), DataType::Utf8) => (
ScalarValue::Utf8(Some(stats.min)),
ScalarValue::Utf8(Some(stats.max)),
),
_ => {
return None;
}
};

Some((min, max))
}
}

/// Column statistics are used to track statistics for a column in a given file.
Expand Down
68 changes: 17 additions & 51 deletions server/src/query/stream_schema_provider.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,7 @@

use std::{any::Any, collections::HashMap, ops::Bound, sync::Arc};

use arrow_schema::{DataType, Schema, SchemaRef, SortOptions};
use arrow_schema::{Schema, SchemaRef, SortOptions};
use bytes::Bytes;
use chrono::{NaiveDateTime, Timelike, Utc};
use datafusion::{
Expand Down Expand Up @@ -236,57 +236,23 @@ fn partitioned_files(
count += num_rows;
}

let mut statistics = vec![];

for field in table_schema.fields() {
let Some(stats) = column_statistics
.get(field.name())
.and_then(|stats| stats.as_ref())
else {
statistics.push(datafusion::common::ColumnStatistics::default());
break;
};

let datatype = field.data_type();

let (min, max) = match (stats, datatype) {
(TypedStatistics::Bool(stats), DataType::Boolean) => (
ScalarValue::Boolean(Some(stats.min)),
ScalarValue::Boolean(Some(stats.max)),
),
(TypedStatistics::Int(stats), DataType::Int32) => (
ScalarValue::Int32(Some(stats.min as i32)),
ScalarValue::Int32(Some(stats.max as i32)),
),
(TypedStatistics::Int(stats), DataType::Int64) => (
ScalarValue::Int64(Some(stats.min)),
ScalarValue::Int64(Some(stats.max)),
),
(TypedStatistics::Float(stats), DataType::Float32) => (
ScalarValue::Float32(Some(stats.min as f32)),
ScalarValue::Float32(Some(stats.max as f32)),
),
(TypedStatistics::Float(stats), DataType::Float64) => (
ScalarValue::Float64(Some(stats.min)),
ScalarValue::Float64(Some(stats.max)),
),
(TypedStatistics::String(stats), DataType::Utf8) => (
ScalarValue::Utf8(Some(stats.min.clone())),
ScalarValue::Utf8(Some(stats.max.clone())),
),
_ => {
statistics.push(datafusion::common::ColumnStatistics::default());
break;
}
};

statistics.push(datafusion::common::ColumnStatistics {
null_count: None,
max_value: Some(max),
min_value: Some(min),
distinct_count: None,
let statistics = table_schema
.fields()
.iter()
.map(|field| {
column_statistics
.get(field.name())
.and_then(|stats| stats.as_ref())
.and_then(|stats| stats.clone().min_max_as_scalar(field.data_type()))
.map(|(min, max)| datafusion::common::ColumnStatistics {
null_count: None,
max_value: Some(max),
min_value: Some(min),
distinct_count: None,
})
.unwrap_or_default()
})
}
.collect();

let statistics = datafusion::common::Statistics {
num_rows: Some(count as usize),
Expand Down