4 ms·
My problem with this is that if you do the same thing, but remove the value 12 from the data you get 4 out of 8 values that are outside 1 stdev...
by pelliphant 6y ago
My problem with this is that if you do the same thing, but remove the value 12 from the data you get 4 out of 8 values that are outside 1 stdev...
- moehm 6y agoYou don't need to and probably shouldn't remove the outliers from your dataset. You can easily query it with: SELECT * FROM series WHERE is_anomaly = f; If you want you can wrap a (materialized) view around this.
- pelliphant 6y agoThnx, but my point was that if there are no outliers, there will still be values that are located more than one stdev from the average.