aboutsummaryrefslogtreecommitdiffstats
path: root/yql/essentials/udfs
diff options
context:
space:
mode:
authorvvvv <vvvv@yandex-team.com>2025-02-28 10:39:13 +0300
committervvvv <vvvv@yandex-team.com>2025-02-28 11:42:27 +0300
commita4ebae0970f4e2748cb954f4fd56b40b42841809 (patch)
treeea7df8ec396ad5916a42577c692c6eefdd6e78bf /yql/essentials/udfs
parentb0a2365a3ba58c8b1c2ef256d2e061662b2b5900 (diff)
downloadydb-a4ebae0970f4e2748cb954f4fd56b40b42841809.tar.gz
YQL-19495 handle NaNs in TDigest
commit_hash:6ceaf9a8cc4d034c2829780bed37396d25f9056d
Diffstat (limited to 'yql/essentials/udfs')
-rw-r--r--yql/essentials/udfs/common/stat/static/stat_udf.h9
-rw-r--r--yql/essentials/udfs/common/stat/test/canondata/result.json7
-rw-r--r--yql/essentials/udfs/common/stat/test/canondata/test.test_nan_/results.txt46
-rw-r--r--yql/essentials/udfs/common/stat/test/cases/nan.sql2
-rw-r--r--yql/essentials/udfs/common/stat/test/ya.make13
-rw-r--r--yql/essentials/udfs/common/stat/ya.make1
6 files changed, 75 insertions, 3 deletions
diff --git a/yql/essentials/udfs/common/stat/static/stat_udf.h b/yql/essentials/udfs/common/stat/static/stat_udf.h
index f0c11a6812..36a1bad7c4 100644
--- a/yql/essentials/udfs/common/stat/static/stat_udf.h
+++ b/yql/essentials/udfs/common/stat/static/stat_udf.h
@@ -22,7 +22,7 @@ namespace {
UdfTerminate((TStringBuilder() << GetPos() << " Invalid combination of delta/K values").data());
}
- return TUnboxedValuePod(new TDigestResource(delta, K, args[0].Get<double>()));
+ return TUnboxedValuePod(new TDigestResource(delta, K, args[0].Get<double>(), true));
}
SIMPLE_STRICT_UDF(TTDigest_AddValue, TResource<DigestResourceName>(TResource<DigestResourceName>, double)) {
@@ -46,14 +46,17 @@ namespace {
SIMPLE_UDF(TTDigest_Deserialize, TResource<DigestResourceName>(char*)) {
Y_UNUSED(valueBuilder);
- return TUnboxedValuePod(new TDigestResource(TString(args[0].AsStringRef())));
+ return TUnboxedValuePod(new TDigestResource(TString(args[0].AsStringRef()), true));
}
SIMPLE_STRICT_UDF(TTDigest_Merge, TResource<DigestResourceName>(TResource<DigestResourceName>, TResource<DigestResourceName>)) {
Y_UNUSED(valueBuilder);
TDigestResource::Validate(args[0]);
TDigestResource::Validate(args[1]);
- return TUnboxedValuePod(new TDigestResource(static_cast<TDigestResource*>(args[0].AsBoxed().Get())->Get(), static_cast<TDigestResource*>(args[1].AsBoxed().Get())->Get()));
+ return TUnboxedValuePod(new TDigestResource(
+ static_cast<TDigestResource*>(args[0].AsBoxed().Get())->Get(),
+ static_cast<TDigestResource*>(args[1].AsBoxed().Get())->Get(),
+ true));
}
/*
diff --git a/yql/essentials/udfs/common/stat/test/canondata/result.json b/yql/essentials/udfs/common/stat/test/canondata/result.json
new file mode 100644
index 0000000000..44314e0309
--- /dev/null
+++ b/yql/essentials/udfs/common/stat/test/canondata/result.json
@@ -0,0 +1,7 @@
+{
+ "test.test[nan]": [
+ {
+ "uri": "file://test.test_nan_/results.txt"
+ }
+ ]
+}
diff --git a/yql/essentials/udfs/common/stat/test/canondata/test.test_nan_/results.txt b/yql/essentials/udfs/common/stat/test/canondata/test.test_nan_/results.txt
new file mode 100644
index 0000000000..1ae063d52a
--- /dev/null
+++ b/yql/essentials/udfs/common/stat/test/canondata/test.test_nan_/results.txt
@@ -0,0 +1,46 @@
+[
+ {
+ "Write" = [
+ {
+ "Type" = [
+ "ListType";
+ [
+ "StructType";
+ [
+ [
+ "column0";
+ [
+ "OptionalType";
+ [
+ "DataType";
+ "Double"
+ ]
+ ]
+ ];
+ [
+ "column1";
+ [
+ "OptionalType";
+ [
+ "DataType";
+ "Double"
+ ]
+ ]
+ ]
+ ]
+ ]
+ ];
+ "Data" = [
+ [
+ [
+ "1.1"
+ ];
+ [
+ "nan"
+ ]
+ ]
+ ]
+ }
+ ]
+ }
+] \ No newline at end of file
diff --git a/yql/essentials/udfs/common/stat/test/cases/nan.sql b/yql/essentials/udfs/common/stat/test/cases/nan.sql
new file mode 100644
index 0000000000..5ab4027b05
--- /dev/null
+++ b/yql/essentials/udfs/common/stat/test/cases/nan.sql
@@ -0,0 +1,2 @@
+select percentile(x,0.99),percentile(x,1.0)
+from (values (double("nan")),(1.1),(0.5)) as a(x)
diff --git a/yql/essentials/udfs/common/stat/test/ya.make b/yql/essentials/udfs/common/stat/test/ya.make
new file mode 100644
index 0000000000..4a14f530f1
--- /dev/null
+++ b/yql/essentials/udfs/common/stat/test/ya.make
@@ -0,0 +1,13 @@
+YQL_UDF_TEST()
+
+DEPENDS(yql/essentials/udfs/common/stat)
+
+TIMEOUT(300)
+
+SIZE(MEDIUM)
+
+IF (SANITIZER_TYPE == "memory")
+ TAG(ya:not_autocheck) # YQL-15385
+ENDIF()
+
+END()
diff --git a/yql/essentials/udfs/common/stat/ya.make b/yql/essentials/udfs/common/stat/ya.make
index 8a5538b371..d1e622b444 100644
--- a/yql/essentials/udfs/common/stat/ya.make
+++ b/yql/essentials/udfs/common/stat/ya.make
@@ -18,6 +18,7 @@ YQL_UDF_CONTRIB(stat_udf)
IF (NOT EXPORT_CMAKE)
RECURSE_FOR_TESTS(
+ test
ut
)
ENDIF()