Partially revert commit 3d3bf62f30200500637b24fdb7b992a99f9704c3.

On reflection, the pre-existing logic in ANALYZE is specifically meant to
compare the frequency of a candidate MCV against the estimated frequency of
a random distinct value across the whole table.  The change to compare it
against the average frequency of values actually seen in the sample doesn't
seem very principled, and if anything it would make us less likely not more
likely to consider a value an MCV.  So revert that, but keep the aspect of
considering only nonnull values, which definitely is correct.

In passing, rename the local variables in these stanzas to
"ndistinct_table", to avoid confusion with the "ndistinct" that appears at
an outer scope in compute_scalar_stats.

diff --git a/src/backend/commands/analyze.c b/src/backend/commands/analyze.c
index 44a4b3ff1e3b7374048ba425a7596f97374420ae..1283de03348aff0ddb3a08d7051721f73dd7760d 100644 (file)
--- a/src/backend/commands/analyze.c
+++ b/src/backend/commands/analyze.c
@@ -2133,13 +2133,15 @@ compute_distinct_stats(VacAttrStatsP stats,
        }
        else
        {
-           /* d here is the same as d in the Haas-Stokes formula */
-           int         d = nonnull_cnt - summultiple + nmultiple;
+           double      ndistinct_table = stats->stadistinct;
            double      avgcount,
                        mincount;
 
+           /* Re-extract estimate of # distinct nonnull values in table */
+           if (ndistinct_table < 0)
+               ndistinct_table = -ndistinct_table * totalrows;
            /* estimate # occurrences in sample of a typical nonnull value */
-           avgcount = (double) nonnull_cnt / (double) d;
+           avgcount = (double) nonnull_cnt / ndistinct_table;
            /* set minimum threshold count to store a value */
            mincount = avgcount * 1.25;
            if (mincount < 2)
@@ -2493,14 +2495,16 @@ compute_scalar_stats(VacAttrStatsP stats,
        }
        else
        {
-           /* d here is the same as d in the Haas-Stokes formula */
-           int         d = ndistinct + toowide_cnt;
+           double      ndistinct_table = stats->stadistinct;
            double      avgcount,
                        mincount,
                        maxmincount;
 
+           /* Re-extract estimate of # distinct nonnull values in table */
+           if (ndistinct_table < 0)
+               ndistinct_table = -ndistinct_table * totalrows;
            /* estimate # occurrences in sample of a typical nonnull value */
-           avgcount = (double) values_cnt / (double) d;
+           avgcount = (double) nonnull_cnt / ndistinct_table;
            /* set minimum threshold count to store a value */
            mincount = avgcount * 1.25;
            if (mincount < 2)