tests: more test corrections

Author: aryan-212

datafusion/core/tests/dataframe/mod.rs

@@ -1206,23 +1206,23 @@ async fn window_using_aggregates() -> Result<()> {
     |             |          |                 |               |        |     |      | 1  | -85  |
     | -85         | -101     | 14              | -12           | -12    | 83  | -101 | 4  | -54  |
     | -85         | -101     | 17              | -25           | -25    | 83  | -101 | 5  | -31  |
-    | -85         | -12      | 10              | -34           | -34    | 83  | -85  | 3  | 13   |
+    | -85         | -12      | 10              | -43           | -34    | 83  | -85  | 3  | 13   |
     | -85         | -25      | 3               | -56           | -56    | -25 | -85  | 1  | -5   |
-    | -85         | -31      | 18              | -28           | -28    | 83  | -101 | 5  | 36   |
+    | -85         | -31      | 18              | -31           | -28    | 83  | -101 | 5  | 36   |
     | -85         | -38      | 16              | -25           | -25    | 83  | -101 | 4  | 65   |
     | -85         | -43      | 7               | -43           | -43    | 83  | -85  | 2  | 45   |
-    | -85         | -48      | 6               | -36           | -36    | 83  | -85  | 2  | -43  |
-    | -85         | -5       | 4               | -40           | -40    | -5  | -85  | 1  | 83   |
-    | -85         | -54      | 15              | -18           | -18    | 83  | -101 | 4  | -38  |
-    | -85         | -56      | 2               | -70           | -70    | -56 | -85  | 1  | -25  |
+    | -85         | -48      | 6               | -48           | -36    | 83  | -85  | 2  | -43  |
+    | -85         | -5       | 4               | -56           | -40    | -5  | -85  | 1  | 83   |
+    | -85         | -54      | 15              | -25           | -18    | 83  | -101 | 4  | -38  |
+    | -85         | -56      | 2               | -85           | -70    | -56 | -85  | 1  | -25  |
     | -85         | -72      | 9               | -43           | -43    | 83  | -85  | 3  | -12  |
     | -85         | -85      | 1               | -85           | -85    | -85 | -85  | 1  | -56  |
-    | -85         | 13       | 11              | -18           | -18    | 83  | -85  | 3  | 14   |
+    | -85         | 13       | 11              | -25           | -18    | 83  | -85  | 3  | 14   |
     | -85         | 13       | 11              | -25           | -25    | 83  | -85  | 3  | 13   |
     | -85         | 14       | 12              | -12           | -12    | 83  | -85  | 3  | 17   |
-    | -85         | 17       | 13              | -8            | -8     | 83  | -85  | 4  | -101 |
-    | -85         | 45       | 8               | -34           | -34    | 83  | -85  | 3  | -72  |
-    | -85         | 65       | 17              | -18           | -18    | 83  | -101 | 5  | -101 |
+    | -85         | 17       | 13              | -12           | -8     | 83  | -85  | 4  | -101 |
+    | -85         | 45       | 8               | -43           | -34    | 83  | -85  | 3  | -72  |
+    | -85         | 65       | 17              | -25           | -18    | 83  | -101 | 5  | -101 |
     | -85         | 83       | 5               | -25           | -25    | 83  | -85  | 2  | -48  |
     +-------------+----------+-----------------+---------------+--------+-----+------+----+------+
     "

datafusion/sqllogictest/test_files/aggregate.slt

@@ -1351,7 +1351,7 @@ DROP TABLE median_window_test;
 query RT
 select approx_median(arrow_cast(col_f32, 'Float16')), arrow_typeof(approx_median(arrow_cast(col_f32, 'Float16'))) from median_table;
 ----
-2.75 Float16
+2.1992188 Float16
 
 # This shouldn't be NaN, see:
 # https://github.com/apache/datafusion/issues/18945
@@ -1369,7 +1369,7 @@ select
   arrow_typeof(approx_percentile_cont(0.5) within group (order by arrow_cast(col_f32, 'Float16')))
 from median_table;
 ----
-2.75 Float16
+2.1992188 Float16
 
 query ?T
 select approx_median(NULL), arrow_typeof(approx_median(NULL)) from median_table;
@@ -1388,7 +1388,7 @@ select median(c), arrow_typeof(median(c)) from t;
 query RT
 select approx_median(c), arrow_typeof(approx_median(c)) from t;
 ----
-0.00035 Float64
+0.0003 Float64
 
 statement ok
 drop table t;
@@ -2322,7 +2322,7 @@ SELECT c1, approx_percentile_cont(0.95) WITHIN GROUP (ORDER BY c3) AS c3_p95 FRO
 a 65
 b 68
 c 118
-d 123
+d 125
 e 112
 
 
@@ -2333,7 +2333,7 @@ SELECT c1, approx_percentile_cont(c3, 0.95) AS c3_p95 FROM aggregate_test_100 GR
 a 65
 b 68
 c 118
-d 123
+d 125
 e 112
 
 
@@ -2344,7 +2344,7 @@ SELECT c1, approx_percentile_cont(c2, 0.95) AS c2, approx_percentile_cont(c3, 0.
 a 5 65
 b 5 68
 c 5 118
-d 5 123
+d 5 125
 e 5 112
 
 # error is unique to this UDAF
@@ -2362,11 +2362,11 @@ e 3 40.333333333333
 query TI
 SELECT c1, approx_percentile_cont(0.95) WITHIN GROUP (ORDER BY c3 DESC) AS c3_p95 FROM aggregate_test_100 GROUP BY 1 ORDER BY 1
 ----
-a -100
-b -111
+a -101
+b -117
 c -107
-d -98
-e -85
+d -99
+e -86
 
 # csv_query_approx_percentile_cont_with_weight (2)
 query TI
@@ -2375,7 +2375,7 @@ SELECT c1, approx_percentile_cont_with_weight(1, 0.95) WITHIN GROUP (ORDER BY c3
 a 65
 b 68
 c 118
-d 123
+d 125
 e 112
 
 # csv_query_approx_percentile_cont_with_weight alternate syntax
@@ -2385,18 +2385,18 @@ SELECT c1, approx_percentile_cont_with_weight(c3, 1, 0.95) AS c3_p95 FROM aggreg
 a 65
 b 68
 c 118
-d 123
+d 125
 e 112
 
 
 query TI
 SELECT c1, approx_percentile_cont_with_weight(1, 0.95) WITHIN GROUP (ORDER BY c3 DESC) AS c3_p95 FROM aggregate_test_100 GROUP BY 1 ORDER BY 1
 ----
-a -100
-b -111
+a -101
+b -117
 c -107
-d -98
-e -85
+d -99
+e -86
 
 # csv_query_approx_percentile_cont_with_histogram_bins
 query TI
@@ -2405,7 +2405,7 @@ SELECT c1, approx_percentile_cont(0.95, 200) WITHIN GROUP (ORDER BY c3) AS c3_p9
 a 65
 b 68
 c 118
-d 123
+d 125
 e 112
 
 query TI
@@ -8804,9 +8804,9 @@ FROM (SELECT * FROM stream_test ORDER BY g LIMIT 10000)
 GROUP BY g
 ORDER BY g;
 ----
-1 2 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5
-2 2 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5
-3 2 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5
+1 2 1 1.5 1.5 1 1 1.5 1 1
+2 2 1 1.5 1.5 1 1 1.5 1 1
+3 2 1 1.5 1.5 1 1 1.5 1 1
 
 # Config reset
 

datafusion/sqllogictest/test_files/aggregate_skip_partial.slt

@@ -321,11 +321,11 @@ SELECT c2, median(c5), median(c11) FROM aggregate_test_100 GROUP BY c2 ORDER BY
 query IIR
 SELECT c2, approx_median(c5), approx_median(c11) FROM aggregate_test_100 GROUP BY c2 ORDER BY c2;
 ----
-1 23971150 0.5922606
-2 -562486880 0.43422085
+1 -335410409 0.5780736
+2 -587831330 0.4279275
 3 240273900 0.40199697
 4 762932956 0.48515016
-5 604973998 0.49842384
+5 586844478 0.44318348
 
 # Test approx_distinct for varchar / int
 query III
@@ -405,8 +405,8 @@ SELECT c2, approx_median(c3), approx_median(c11) FROM aggregate_test_100_null GR
 1 12 0.6067944
 2 1 0.46076488
 3 14 0.40154034
-4 -17 0.48515016
-5 -35 0.5536642
+4 -38 0.48515016
+5 -40 0.5536642
 
 # Test approx_distinct with nullable fields
 query II
@@ -525,8 +525,8 @@ FROM aggregate_test_100 GROUP BY c2 ORDER BY c2;
 ----
 1 57 -56
 2 52 -60
-3 71 -74
-4 65 -69
+3 71 -76
+4 65 -79
 5 64 -59
 
 # Test count with nullable fields and filter
@@ -669,22 +669,22 @@ SELECT c2,
 FROM aggregate_test_100_null GROUP BY c2 ORDER BY c2;
 ----
 1 -5 0.6623719
-2 15 0.52930677
+2 1 0.52930677
 3 13 0.32792538
 4 -38 0.49774808
-5 -18 0.49842384
+5 -31 0.44318348
 
 # Test approx_median with nullable fields and nullable filter
 query II
 SELECT c2,
        approx_median(c3) FILTER (WHERE c11 > 0.5)
 FROM aggregate_test_100_null GROUP BY c2 ORDER BY c2;
 ----
-1 33
+1 29
 2 -29
 3 22
 4 -90
-5 -22
+5 -40
 
 statement ok
 DROP TABLE aggregate_test_100_null;

datafusion/sqllogictest/test_files/metadata.slt

@@ -86,7 +86,7 @@ select count(distinct name) from table_with_metadata;
 query I
 select approx_median(distinct id) from table_with_metadata;
 ----
-2
+1
 
 # Regression test: prevent field metadata loss per https://github.com/apache/datafusion/issues/12687
 statement ok