Data Outliers

Contents[Hide]

The Data Outlier functions identify the outliers in a set of input values. You can hide or replace the outliers to display only non-outliers, or the result can include only the outliers to display them or to format them using states.

Identify data outliers
Identify data outliers

The Outlier Interquartile function uses a multiple of the interquartile range (IQR) to determine which values are outliers, typically with a multiple of 1.5. The interquartile range is the difference between the 75th and 25th percentiles. A multiple of 3 is sometimes used to further identify extreme or 'major' outliers, and you can use the parameters of this function to show or hide major and minor outliers separately if you choose.

The Outlier Tau function uses the modified Thompson Tau test to determine the outliers.

1. Syntax

Outlier Interquartile:

OUTIQ(d0,s0,s1,s2,s3,s4,s5,Alignments)

Outlier Tau:

OUTTAU(d0,s0,s1,s2,Alignments)

2. Input

  • d0 - The set of data values for which the outliers is calculated.

3. OUTIQ parameters

  • s0 - The value used to determine the position of the minor outliers. The default value is 1.5 and the supported range is from 1 to 5.
  • s1 - The value used to determine the position of the major outliers. The default value is 3 and the supported range is from 1 to 10.
  • s2 - Indicates that all non-outlier data points should be hidden in the result. The default value is False.
  • s3 - Indicates that all minor outliers should be hidden in the result. The default value is True.
  • s4 - Indicates that all major outlier data points should be hidden in the result. The default value is True.
  • s5 - Indicates what value should be used to replace any hidden outliers. Available choices include:
    • 0 - null values
    • 1 - zero
    • 2 - neighbor's average
    • 3 - global average
  • Alignments (optional) - A hierarchy placeholder to be used as the alignment axis.

4. OUTTAU parameters

  • s0 - Indicates that all non-outlier data points should be hidden in the result. The default value is False.
  • s1 - Indicates that all outliers should be hidden in the result. The default value is True.
  • s2 - Indicates what value should be used to replace any hidden outliers. Available choices include:
    • 0 - null values
    • 1 - zero
    • 2 - neighbor's average
    • 3 - global average
  • Alignments (optional) - A hierarchy placeholder to be used as the alignment axis.

5. Output

The Data Outlier functions generate the following output:

  • The Data Outliers result set, consisting of the input values with non-outlier values hidden and/or with outlier values hidden or replaced as specified in the parameters.

6. See also

Dundas Data Visualization, Inc.
400-15 Gervais Drive
Toronto, ON, Canada
M3C 1Y8

North America: 1.800.463.1492
International: 1.416.467.5100

Dundas Support Hours:
Phone: 9am-6pm, ET, Mon-Fri
Email: 7am-6pm, ET, Mon-Fri