cartodb4/crankshaft
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

adds symmetric option for stddev outlier

Browse Source
This commit is contained in:
Andy Eschbacher 2016-09-29 11:09:10 -04:00
parent 8bc6f69a1b
commit 5754087140
1 changed files with 11 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/pg/sql/18_outliers.sql
View File

@ -12,7 +12,7 @@ END;
$$ LANGUAGE plpgsql; $$ LANGUAGE plpgsql;
-- Find outliers by a percentage above the threshold -- Find outliers by a percentage above the threshold
-- TODO: add symmetric option? `symmetric boolean DEFAULT false` -- TODO: add symmetric option? `is_symmetric boolean DEFAULT false`
CREATE OR REPLACE FUNCTION CDB_PercentOutlier(column_values numeric[], outlier_fraction numeric, ids int[]) CREATE OR REPLACE FUNCTION CDB_PercentOutlier(column_values numeric[], outlier_fraction numeric, ids int[])
RETURNS TABLE(outlier boolean, rowid int) RETURNS TABLE(outlier boolean, rowid int)
@ -42,7 +42,7 @@ $$ LANGUAGE plpgsql;
-- Find outliers above a given number of standard deviations from the mean -- Find outliers above a given number of standard deviations from the mean
CREATE OR REPLACE FUNCTION CDB_StdDevOutlier(attrs numeric[], num_deviations numeric, ids int[]) CREATE OR REPLACE FUNCTION CDB_StdDevOutlier(attrs numeric[], num_deviations numeric, ids int[], is_symmetric boolean DEFAULT true)
RETURNS TABLE(outlier boolean, rowid int) RETURNS TABLE(outlier boolean, rowid int)
AS $$ AS $$
DECLARE DECLARE
@ -58,10 +58,15 @@ BEGIN
RAISE EXCEPTION 'Standard deviation of input data is zero'; RAISE EXCEPTION 'Standard deviation of input data is zero';
END IF; END IF;
IF is_symmetric THEN
SELECT array_agg( SELECT array_agg(
abs(i - avg_val) / stddev_val > num_deviations) INTO out_vals abs(i - avg_val) / stddev_val > num_deviations) INTO out_vals
FROM unnest(attrs) As x(i); FROM unnest(attrs) As x(i);
ELSE
SELECT array_agg(
(i - avg_val) / stddev_val > num_deviations) INTO out_vals
FROM unnest(attrs) As x(i);
END IF;
RETURN QUERY RETURN QUERY
SELECT unnest(out_vals) As outlier, SELECT unnest(out_vals) As outlier,