add test

zhyass · zhyass · commit 8f7a54ae0627 · 2025-09-25T00:09:32.000+08:00
diff --git a/src/query/catalog/src/statistics/basic_statistics.rs b/src/query/catalog/src/statistics/basic_statistics.rs
@@ -131,6 +131,7 @@ impl BasicColumnStatistics {
         })
     }
 
+    // Inspired by duckdb (https://github.com/duckdb/duckdb/blob/main/src/storage/statistics/distinct_statistics.cpp#L55-L69)
     fn estimate_ndv(ndv: u64, stats_row_count: u64, num_rows: u64) -> u64 {
         if stats_row_count == 0 || ndv == 0 {
             return num_rows;
@@ -161,5 +162,6 @@ mod tests {
         assert_eq!(BasicColumnStatistics::estimate_ndv(0, 1, 3), 3);
         assert_eq!(BasicColumnStatistics::estimate_ndv(1, 1, 3), 3);
         assert_eq!(BasicColumnStatistics::estimate_ndv(12, 100, 3000), 17);
+        assert_eq!(BasicColumnStatistics::estimate_ndv(6000, 10000, 1000000), 219840);
     }
 }

Original file line number	Diff line number	Diff line change
`@@ -131,6 +131,7 @@ impl BasicColumnStatistics {`
`131`	`131`	`})`
`132`	`132`	`}`
`133`	`133`
	`134`	`+ // Inspired by duckdb (https://github.com/duckdb/duckdb/blob/main/src/storage/statistics/distinct_statistics.cpp#L55-L69)`
`134`	`135`	`fn estimate_ndv(ndv: u64, stats_row_count: u64, num_rows: u64) -> u64 {`
`135`	`136`	`if stats_row_count == 0 \|\| ndv == 0 {`
`136`	`137`	`return num_rows;`
`@@ -161,5 +162,6 @@ mod tests {`
`161`	`162`	`assert_eq!(BasicColumnStatistics::estimate_ndv(0, 1, 3), 3);`
`162`	`163`	`assert_eq!(BasicColumnStatistics::estimate_ndv(1, 1, 3), 3);`
`163`	`164`	`assert_eq!(BasicColumnStatistics::estimate_ndv(12, 100, 3000), 17);`
	`165`	`+ assert_eq!(BasicColumnStatistics::estimate_ndv(6000, 10000, 1000000), 219840);`
`164`	`166`	`}`
`165`	`167`	`}`