Statistics

The Statistics tab show the statistics calculated by the Explore Data node.

Attributes are listed the Statistics grid. For each attribute, the name, data type, histogram, and summary of the statistics are displayed.

To view a large version of the histogram for an attribute, select the attribute. The histogram is displayed below the grid. If the attribute has null values, then the histogram has a separate bin labeled Null bin. The histogram may also contain an Others bin consisting of values that are not NULL but are not in any other bin.

For more information about the histogram, go to the large histogram and right-click the attribute name. These choices are available:

• Copy to Clipboard: Copies the histogram to the Microsoft Windows clipboard. You can paste this histogram in to any rich editor, such as a word processor or image editor.

• Save Image As: Exports the image to a PNG file that can be stored in the file system.

• View Data: Displays the data used to create the histogram in a pop up window. You can search for attribute name, attribute value, or attribute percent. Click Close when you have finished.

Histograms created by Oracle Data Miner use a sample of the data and can be slightly different each time they are displayed.

For all data types, a histogram is created. The number of distinct values, the percentage of NULL values, and the number of distinct values are displayed.

Additional statistics calculated depend on the data type of the attribute:

• For character attributes VARCHAR2, the Mode is calculated.

• For numeric attributes NUMBER or FLOAT,the following are calculated:

  • Average

  • Minimum value

  • Maximum value

  • Standard deviation

  • Skewness (a measure of the asymmetry of the distribution)

  • Kurtosis (a measure of flatness or peakedness of the curve describing a frequency distribution in the region about its mode)