Order mozfun.glam.histogram_fill_buckets by string sorting (#1497)

* Order fill buckets by string sorting

* Add buckets in the correct order

* Simplify histogram_fill_buckets into single statement

* Remove extra casting

sql/mozfun/glam/histogram_fill_buckets/udf.sql

@@ -5,26 +5,18 @@ CREATE OR REPLACE FUNCTION glam.histogram_fill_buckets(
 )
 RETURNS ARRAY<STRUCT<key STRING, value FLOAT64>> AS (
   -- Given a MAP `input_map`, fill in any missing keys with value `0.0`
-  (
-    WITH total_counts AS (
-      SELECT
-        key,
-        COALESCE(e.value, 0.0) AS value
-      FROM
-        UNNEST(buckets) AS key
-      LEFT JOIN
-        UNNEST(input_map) AS e
-      ON
-        SAFE_CAST(key AS STRING) = e.key
-    )
-    SELECT
-      ARRAY_AGG(
-        STRUCT<key STRING, value FLOAT64>(SAFE_CAST(key AS STRING), value)
-        ORDER BY
-          CAST(key AS int64)
-      )
+  ARRAY(
+    SELECT AS STRUCT
+      key,
+      COALESCE(e.value, 0.0) AS value
     FROM
-      total_counts
+      UNNEST(buckets) AS key
+    LEFT JOIN
+      UNNEST(input_map) AS e
+    ON
+      key = e.key
+    ORDER BY
+      key
   )
 );
 
@@ -45,9 +37,9 @@ SELECT
       ["0"]
     )
   ),
-  -- return ordered keys
+  -- return ordered keys by string value
   assert.array_equals(
-    ARRAY<STRUCT<key STRING, value FLOAT64>>[("2", 1.0), ("11", 0.0)],
+    ARRAY<STRUCT<key STRING, value FLOAT64>>[("11", 0.0), ("2", 1.0)],
     glam.histogram_fill_buckets(
       ARRAY<STRUCT<key STRING, value FLOAT64>>[("0", 1.0), ("2", 1.0)],
       ["11", "2"]

sql/mozfun/glam/histogram_fill_buckets_dirichlet/udf.sql

@@ -6,33 +6,25 @@ CREATE OR REPLACE FUNCTION glam.histogram_fill_buckets_dirichlet(
 )
 RETURNS ARRAY<STRUCT<key STRING, value FLOAT64>> AS (
   -- Given a MAP `input_map`, fill in any missing keys with value `0.0`
-  (
-    WITH total_counts AS (
-      SELECT
-        key,
-        -- Dirichlet distribution density for each bucket in a histogram.
-        -- Given {k1: p1,k2:p2} where p’s are proportions(and p1, p2 sum to 1)
-        -- return {k1: (P1+1/K) / (nreporting+1), k2:(P2+1/K) / (nreporting+1)}.
-        -- https://docs.google.com/document/d/1ipy1oFIKDvHr3R6Ku0goRjS11R1ZH1z2gygOGkSdqUg
-        SAFE_DIVIDE(
-          COALESCE(e.value, 0.0) + SAFE_DIVIDE(1, ARRAY_LENGTH(buckets)),
-          total_users + 1
-        ) AS value
-      FROM
-        UNNEST(buckets) AS key
-      LEFT JOIN
-        UNNEST(input_map) AS e
-      ON
-        SAFE_CAST(key AS STRING) = e.key
-    )
-    SELECT
-      ARRAY_AGG(
-        STRUCT<key STRING, value FLOAT64>(SAFE_CAST(key AS STRING), value)
-        ORDER BY
-          CAST(key AS INT64)
-      )
+  ARRAY(
+    SELECT AS STRUCT
+      key,
+      -- Dirichlet distribution density for each bucket in a histogram.
+      -- Given {k1: p1,k2:p2} where p’s are proportions(and p1, p2 sum to 1)
+      -- return {k1: (P1+1/K) / (nreporting+1), k2:(P2+1/K) / (nreporting+1)}.
+      -- https://docs.google.com/document/d/1ipy1oFIKDvHr3R6Ku0goRjS11R1ZH1z2gygOGkSdqUg
+      SAFE_DIVIDE(
+        COALESCE(e.value, 0.0) + SAFE_DIVIDE(1, ARRAY_LENGTH(buckets)),
+        total_users + 1
+      ) AS value
     FROM
-      total_counts
+      UNNEST(buckets) AS key
+    LEFT JOIN
+      UNNEST(input_map) AS e
+    ON
+      key = e.key
+    ORDER BY
+      key
   )
 );
 
@@ -59,9 +51,9 @@ SELECT
       2
     )
   ),
-  -- out of order keys
+  -- keys may not non-integer values, so we must make do with string ordering
   assert.array_equals(
-    ARRAY<STRUCT<key STRING, value FLOAT64>>[("2", (1 + (1 / 2)) / 3), ("11", (1 + (1 / 2)) / 3)],
+    ARRAY<STRUCT<key STRING, value FLOAT64>>[("11", (1 + (1 / 2)) / 3), ("2", (1 + (1 / 2)) / 3)],
     glam.histogram_fill_buckets_dirichlet(
       ARRAY<STRUCT<key STRING, value FLOAT64>>[("11", 1.0), ("2", 1.0)],
       ["11", "2"],