diff options
author | Tom Lane | 2009-01-06 23:46:06 +0000 |
---|---|---|
committer | Tom Lane | 2009-01-06 23:46:06 +0000 |
commit | 01e050fd2bcbef383911f1ead4c0f3be275e2d08 (patch) | |
tree | 1830b63cd83285b69f49dce6ad5d0052920ef630 | |
parent | 4c3d441e090b18095ef8e2bf8d35e89f8b558dc4 (diff) |
Clarify a confusing comment about MCVs vs histogram entries.
Per Nathan Boley.
-rw-r--r-- | src/backend/commands/analyze.c | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/src/backend/commands/analyze.c b/src/backend/commands/analyze.c index 7fa8cee2e1..311397b360 100644 --- a/src/backend/commands/analyze.c +++ b/src/backend/commands/analyze.c @@ -2149,7 +2149,9 @@ compute_scalar_stats(VacAttrStatsP stats, * least 2 instances in the sample. Also, we won't suppress values * that have a frequency of at least 1/K where K is the intended * number of histogram bins; such values might otherwise cause us to - * emit duplicate histogram bin boundaries. + * emit duplicate histogram bin boundaries. (We might end up with + * duplicate histogram entries anyway, if the distribution is skewed; + * but we prefer to treat such values as MCVs if at all possible.) */ if (track_cnt == ndistinct && toowide_cnt == 0 && stats->stadistinct > 0 && |