Table of Contents

• Introduction
  • Purpose
  • Description
• Using The Analytics Package Within NetCharts Designer
  • General Use
  • A Basic Example
• General Purpose Functions
  • Aggregate Column
  • Grade Column
  • Delete Column(s)
  • Delete Row(s)
  • Include Headers
  • Count Group Names
  • Sum Group Names
  • Sort 1 Column
  • Two Column Operation
• Statistical Functions
  • Bin Data
  • Univariate Metrics
  • Histogram Distribution
  • Pareto Analysis
  • Normal Distribution
  • Quantile
  • Mode
• Calculus Functions
  • Area
  • Differentiation
  • Curve Length
  • Surface Metrics
  • Regression
  • Fourier
  • Filter
  • Track Order
  • Interpolation
• Financial Functions
  • Moving Average
  • Bollinger Bands
  • Linear Regression
  • Stochastic Oscillator
  • Chaikin Oscillator
  • Ultimate Oscillator
  • Price Oscillator
• Statistical Process Control (SPC) Analytics
  • SPC Run Rules
  • SPC Attributes Chart – c
  • SPC Attributes Chart – np
  • SPC Attributes Chart – p
  • SPC Attributes Chart – u
  • SPC Variables Chart – EWMA
  • SPC Variables Chart – Histogram
  • SPC Variables Chart – Median-R
  • SPC Variables Chart – Median-s
  • SPC Variables Chart – Probability
  • SPC Variables Chart – XBar-R
  • SPC Variables Chart – XBar-s
• Appendix
  • Regression
  • Fourier, Filter and Track Order
  • Univariate Metrics
• Where can I go for additional resources and help?

Introduction

This document provides a comprehensive description of the content, methodology, and features of the Analytics Library built into NetCharts Server and Designer.

Purpose

Analytic packages are used to extract useful information from raw data by applying various algorithms to the original data set. For example, a grocery store manager could apply a Pareto function to raw register receipts to see a ranking of the most popular products in the store. Trend and forecasting lines can be applied to gauge future sales. Statistical Process Control (SPC) functions can determine if the store is meeting sales and/or quality expectations.

This Analytics Library provides the ability for users to apply sophisticated data transformations to raw or minimally processed data. The resultant data can then be easily displayed in graphical or tabular formats.

Description

The Analytics Library described in this document is implemented in the NetCharts Server and Designer as process modules. The process module interface allows data to be modified thru one or more transformations. In NetCharts Designer, the process modules may be selected via the Functions tab on the NDX data editor.

All modules are described in detail in below. Unit tests with sample data are provided in order to illustrate the operation of each module (download the unit test project).

In addition to module-specific parameters, each module also supports an Enable Processing function. This allows the module to be turned on or off to simplify debugging.

The Appendix provides additional information about Regression, Fourier analysis and Univariate Metrics.

Using The Analytics Package Within NetCharts Designer

General Use

Follow these general steps to use the Analytics package with NetCharts Designer:

1. Create a new project and an NDX that accesses the raw data.
2. Apply the Analytics functions to the NDX using the Process Modules interface.
3. Map the resultant data into a graph and customize as needed.

A Basic Example

This example will show the basic use of the NetCharts Analytics package from within NetCharts Designer. We will create a simple SPC run chart. This example assumes general familiarity with the NetCharts Designer product.

Create a new project and an NDX that accesses the raw data

• Create a new project called AnalyticsTest
• Create a new NDX file called data that uses the Hypersonic data source. Use the following SQL:

SELECT MONTH, “Server 1 Hits” FROM PAGEHITTABLE

• Select Finish. Set the Date Format to MMM-yy (ex. Jan-08) in the editor.
   
  
   

Apply the Analytics functions to the NDX

• Select the Functions tab at the bottom of the editor.
• In the Process Modules section, select Add.
• Expand the Visual Mining Analytics branch and scroll to, and select, SPC Variables Chart Run
   
  
   
• Select OK
• Enter in the following Information (Note that the leftmost column’s index is ‘0’)
  • Column Index For Sample Id: 0
  • Column Index For Item Value: 1
  • Include Specifications: No
  • LSL:
  • USL:
  • Target:
  • Piecewise: No
  • Column Index For Piecewise:
  • Piecewise Range:
  • Result Type: Chart Data
  • Include Original Data: Yes
  • Enable Processing: Yes

• Select OK. The new data is written to the output.
   
  
   

Map the resultant data into a graph

• Create a new line graph
• Go to the Data Source Binding panel
• Select Lines from the tree control.
• Select data.ndx from the Data Source control
• Select the 2nd, 4th, 5th, and 6th columns. (This is the data, the XBAR, and +/- 3 sigma).
• Select the Select Rows button. This will automatically select all but the first row and the graph will populate with data.
   
  
   
• Select Line Axis Labels from the tree control.
• Select the data.ndx from the Data Source control
• Set the menu over the first column to Axis Labels.
• Select the Select Rows button. The axis labels will populate.
• Select Legend Items from the tree control.
• Select data.ndx from the Data Source control.
• Select the 2nd, 4th, 5th, and 6th columns
• Select Row 0. The legend labels will populate.

At this point, enable the legend by selecting the Presentation tab then the Legend tree. Here check the Show Legend check box and place the legend at the top, horizontally.
 

 
Finally, use the standard NetCharts Designer features can be used to customize the look of the chart and create pages for deployment.