D267: statistics aggregates, ADD_MONTHS, MONTHS_BETWEEN and NVL2 from receipts

cases/translation/functions/add-months.toml

@@ -86,3 +86,18 @@ rows = [
 query = "SELECT ADD_MONTHS('2024-01-15'::DATE, n) AS r FROM input ORDER BY ordinal"
 order = "ordered"
 tags = ["boundary"]
+
+[[case]]
+id = "t008_negative_and_day_31"
+description = "Negative counts from month ends and day 31 into shorter months, across a year. Catches: the month end lost going backwards, or day 31 overflowing into the next month"
+columns = [{ name = "d", type = "DATE" }]
+rows = [
+  ["'2024-03-31'"],
+  ["'2024-01-31'"],
+  ["'2024-05-31'"],
+  ["'2023-01-29'"],
+  ["'2024-12-15'"],
+]
+query = "SELECT ADD_MONTHS(d, -1) AS m1, ADD_MONTHS(d, 1) AS p1, ADD_MONTHS(d, -13) AS m13 FROM input ORDER BY ordinal"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/corr.corr.toml

@@ -83,3 +83,30 @@ rows = [
 query = "SELECT CORR(y, x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Pairs near 1e9 with small spreads. Catches: a sum-of-products formula losing the digits of the spread"
+columns = [{ name = "y", type = "NUMBER(38,0)" }, { name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000002", "1000000001"],
+  ["1000000004", "1000000002"],
+  ["1000000005", "1000000003"],
+  ["1000000009", "1000000004"],
+]
+query = "SELECT CORR(y, x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "NUMBER(10,2) columns. Catches: the result type or scale of decimal inputs"
+columns = [{ name = "y", type = "NUMBER(10,2)" }, { name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25", "2.50"],
+  ["2.50", "2.75"],
+  ["4.10", "3.00"],
+]
+query = "SELECT CORR(y, x) AS r, SYSTEM$TYPEOF(CORR(y, x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/covar-pop.covar-pop.toml

@@ -70,3 +70,31 @@ rows = [
 query = "SELECT COVAR_POP(y, x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t006_large_mean"
+description = "Pairs near 1e9 with small spreads. Catches: a sum-of-products formula losing the digits of the spread"
+columns = [{ name = "y", type = "NUMBER(38,0)" }, { name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000002", "1000000001"],
+  ["1000000004", "1000000002"],
+  ["1000000005", "1000000003"],
+  ["1000000009", "1000000004"],
+]
+query = "SELECT COVAR_POP(y, x) AS r FROM input"
+order = "ordered"
+expected_divergence = "float_sum_order"
+tags = ["type-combination"]
+
+[[case]]
+id = "t007_decimal"
+description = "NUMBER(10,2) columns. Catches: the result type or scale of decimal inputs"
+columns = [{ name = "y", type = "NUMBER(10,2)" }, { name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25", "2.50"],
+  ["2.50", "2.75"],
+  ["4.10", "3.00"],
+]
+query = "SELECT COVAR_POP(y, x) AS r, SYSTEM$TYPEOF(COVAR_POP(y, x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/covar-samp.covar-samp.toml

@@ -70,3 +70,31 @@ rows = [
 query = "SELECT COVAR_SAMP(y, x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t006_large_mean"
+description = "Pairs near 1e9 with small spreads. Catches: a sum-of-products formula losing the digits of the spread"
+columns = [{ name = "y", type = "NUMBER(38,0)" }, { name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000002", "1000000001"],
+  ["1000000004", "1000000002"],
+  ["1000000005", "1000000003"],
+  ["1000000009", "1000000004"],
+]
+query = "SELECT COVAR_SAMP(y, x) AS r FROM input"
+order = "ordered"
+expected_divergence = "float_sum_order"
+tags = ["type-combination"]
+
+[[case]]
+id = "t007_decimal"
+description = "NUMBER(10,2) columns. Catches: the result type or scale of decimal inputs"
+columns = [{ name = "y", type = "NUMBER(10,2)" }, { name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25", "2.50"],
+  ["2.50", "2.75"],
+  ["4.10", "3.00"],
+]
+query = "SELECT COVAR_SAMP(y, x) AS r, SYSTEM$TYPEOF(COVAR_SAMP(y, x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/months-between.toml

@@ -84,3 +84,29 @@ rows = [
 query = "SELECT MONTHS_BETWEEN(a, b) AS r FROM input ORDER BY ordinal"
 order = "ordered"
 tags = ["boundary"]
+
+[[case]]
+id = "t008_month_ends_with_time"
+description = "Timestamps on month ends and on equal days, with times of day. Catches: the time counted when both are month ends or share a day, or the fraction taken over the month length instead of 31 days"
+columns = [{ name = "a", type = "TIMESTAMP_NTZ" }, { name = "b", type = "TIMESTAMP_NTZ" }]
+rows = [
+  ["'2019-03-31 23:00:00'", "'2019-02-28 01:00:00'"],
+  ["'2024-04-30 00:00:00'", "'2024-03-30 12:00:00'"],
+  ["'2024-01-31 06:00:00'", "'2024-02-29 18:00:00'"],
+  ["'2024-01-10 06:00:00'", "'2024-02-20 18:00:00'"],
+]
+query = "SELECT MONTHS_BETWEEN(a, b) AS r FROM input ORDER BY ordinal"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t009_long_negative"
+description = "A span of years, negative. Catches: the whole months or the fraction miscounted far apart"
+columns = [{ name = "a", type = "DATE" }, { name = "b", type = "DATE" }]
+rows = [
+  ["'2000-01-15'", "'2024-07-20'"],
+  ["'2024-07-20'", "'1999-12-31'"],
+]
+query = "SELECT MONTHS_BETWEEN(a, b) AS r FROM input ORDER BY ordinal"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/stddev-pop.stddev-pop.toml

@@ -104,3 +104,30 @@ rows = [
 query = "SELECT STDDEV_POP(x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Values near 1e9 with a spread of 1. Catches: a sum-of-squares formula losing the digits of the spread"
+columns = [{ name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000001"],
+  ["1000000002"],
+  ["1000000003"],
+  ["1000000004"],
+]
+query = "SELECT STDDEV_POP(x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "A NUMBER(10,2) column. Catches: the result type or scale of a decimal input"
+columns = [{ name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25"],
+  ["2.50"],
+  ["4.10"],
+]
+query = "SELECT STDDEV_POP(x) AS r, SYSTEM$TYPEOF(STDDEV_POP(x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/stddev.stddev-samp.toml

@@ -104,3 +104,30 @@ rows = [
 query = "SELECT STDDEV_SAMP(x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Values near 1e9 with a spread of 1. Catches: a sum-of-squares formula losing the digits of the spread"
+columns = [{ name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000001"],
+  ["1000000002"],
+  ["1000000003"],
+  ["1000000004"],
+]
+query = "SELECT STDDEV_SAMP(x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "A NUMBER(10,2) column. Catches: the result type or scale of a decimal input"
+columns = [{ name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25"],
+  ["2.50"],
+  ["4.10"],
+]
+query = "SELECT STDDEV_SAMP(x) AS r, SYSTEM$TYPEOF(STDDEV_SAMP(x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/stddev.stddev.toml

@@ -104,3 +104,30 @@ rows = [
 query = "SELECT STDDEV(x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Values near 1e9 with a spread of 1. Catches: a sum-of-squares formula losing the digits of the spread"
+columns = [{ name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000001"],
+  ["1000000002"],
+  ["1000000003"],
+  ["1000000004"],
+]
+query = "SELECT STDDEV(x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "A NUMBER(10,2) column. Catches: the result type or scale of a decimal input"
+columns = [{ name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25"],
+  ["2.50"],
+  ["4.10"],
+]
+query = "SELECT STDDEV(x) AS r, SYSTEM$TYPEOF(STDDEV(x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/var-pop.var-pop.toml

@@ -104,3 +104,30 @@ rows = [
 query = "SELECT VAR_POP(x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Values near 1e9 with a spread of 1. Catches: a sum-of-squares formula losing the digits of the spread"
+columns = [{ name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000001"],
+  ["1000000002"],
+  ["1000000003"],
+  ["1000000004"],
+]
+query = "SELECT VAR_POP(x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "A NUMBER(10,2) column. Catches: the result type or scale of a decimal input"
+columns = [{ name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25"],
+  ["2.50"],
+  ["4.10"],
+]
+query = "SELECT VAR_POP(x) AS r, SYSTEM$TYPEOF(VAR_POP(x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/var-samp.var-samp.toml

@@ -104,3 +104,30 @@ rows = [
 query = "SELECT VAR_SAMP(x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Values near 1e9 with a spread of 1. Catches: a sum-of-squares formula losing the digits of the spread"
+columns = [{ name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000001"],
+  ["1000000002"],
+  ["1000000003"],
+  ["1000000004"],
+]
+query = "SELECT VAR_SAMP(x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "A NUMBER(10,2) column. Catches: the result type or scale of a decimal input"
+columns = [{ name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25"],
+  ["2.50"],
+  ["4.10"],
+]
+query = "SELECT VAR_SAMP(x) AS r, SYSTEM$TYPEOF(VAR_SAMP(x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]

cases/translation/functions/variance-pop.variance-pop.toml

@@ -104,3 +104,30 @@ rows = [
 query = "SELECT VARIANCE_POP(x) AS r FROM input WHERE ordinal < 0"
 order = "ordered"
 tags = ["empty-input"]
+
+[[case]]
+id = "t007_large_mean"
+description = "Values near 1e9 with a spread of 1. Catches: a sum-of-squares formula losing the digits of the spread"
+columns = [{ name = "x", type = "NUMBER(38,0)" }]
+rows = [
+  ["1000000001"],
+  ["1000000002"],
+  ["1000000003"],
+  ["1000000004"],
+]
+query = "SELECT VARIANCE_POP(x) AS r FROM input"
+order = "ordered"
+tags = ["type-combination"]
+
+[[case]]
+id = "t008_decimal"
+description = "A NUMBER(10,2) column. Catches: the result type or scale of a decimal input"
+columns = [{ name = "x", type = "NUMBER(10,2)" }]
+rows = [
+  ["1.25"],
+  ["2.50"],
+  ["4.10"],
+]
+query = "SELECT VARIANCE_POP(x) AS r, SYSTEM$TYPEOF(VARIANCE_POP(x)) AS t FROM input"
+order = "ordered"
+tags = ["type-combination"]