Overview

Statistics is the study of how to acquire meaningful information by analyzing data (e.g. a list of numbers). Dundas Chart for ASP.NET provides a complete set of statistical analysis functions, which are implemented using the Statistics class (exposed as the Statistics property of the DataManipulator class).

This Statistics class can be organized into 4 general groups:

• Statistical Tests.
• Statistical Distributions.
• Basic Statistical Functions.
• Utility functions.

It is important to note that all statistical methods that use one or more Series for their input data cannot use indexed data points.

Since these statistical methods cannot use indexed data points, therefore you must:

• Set the XValue property of all data points explicitly.
• Be certain that each series' XValueIndexed property is false.

Statistical Tests

Statistical tests are used for Hypothesis testing. There are 6 different tests: ANOVA, F-Test, Z-Test, T-Tests (Equal Variance, Unequal Variance and Paired). All tests use series and various parameters for input, while the output value for each test is a test result class. These classes have read-only properties, and store the results from the tests.

Unlike the other tests that must have exactly two input series, the Anova test can have more then two input series.

Example

The following sample performs an Anova test for three series of data.

Visual Basic	Copy Code
Imports Dundas.Charting.WebControl ... ' Perform the Anova test. Dim result As AnovaResult = Chart1.DataManipulator.Statistics.Anova(0.6, "Group1,Group2,Group3")

C#	Copy Code
using Dundas.Charting.WebControl; ... // Perform the Anova test. AnovaResult result = Chart1.DataManipulator.Statistics.Anova(0.6, "Group1,Group2,Group3");

Example

The following sample performs a T Test assuming equal variances. The first parameter is the Hypothesis Mean Difference, and in this case the hypothesis is that the first group has a mean which is less the second group mean by 0.6. The second parameter is the probability that the hypothesis test is rejected.

Visual Basic	Copy Code
Imports Dundas.Charting.WebControl ... ' Perform t-test. Dim result As TTestResult = Chart1.DataManipulator.Statistics.TTestEqualVariances(0.6, 0.95, "FirstGroup", "SecondGroup")

C#	Copy Code
using Dundas.Charting.WebControl; ... // Perform t-test. TTestResult result = Chart1.DataManipulator.Statistics.TTestEqualVariances(0.6, 0.95, "FirstGroup", "SecondGroup");

Distribution Formulas

Dundas chart supports 3 Distributions: Normal Distribution, T-Distribution (or Students Distribution), and F-Distribution.

Figure 1: Graph of Standard Normal Distribution (left), T distribution (middle), and F Distribution (right).

The Distribution functions always have one double value for input, and return the probability for that distribution. The Inverse Distribution functions use probability as an input value, along with one, or two degrees of freedom.

Example

This example demonstrates how to calculate the inverse F distribution.

Visual Basic	Copy Code
Imports Dundas.Charting.WebControl ... ' Get the fValue, using probability of .05, and first and second ' degrees of freedom of 3 and 4 respectively. Dim fValue As Double = Chart1.DataManipulator.Statistics.InverseFDistribution(.05, 3, 4)

C#	Copy Code
using Dundas.Charting.WebControl; ... // Get the fValue, using probability of .05, and first and second // degrees of freedom of 3 and 4 respectively. double fValue = Chart1.DataManipulator.Statistics.InverseFDistribution(.05, 3, 4);

Basic Statistical Formulas

The basic statistical functions are: Mean, Median, Variance, Correlation and Covariance. These functions always return a double value, and use one or two series for input.

Example

This example demonstrates how to calculate the Covariance.

Visual Basic	Copy Code
Imports Dundas.Charting.WebControl ... ' Calculating the Covariance. Dim covariance As Double = Chart1.DataManipulator.Statistics.Covariance("FirstGroup", "SecondGroup")

C#	Copy Code
using Dundas.Charting.WebControl; ... // Calculating the Covariance. double covariance = Chart1.DataManipulator.Statistics.Covariance("FirstGroup", "SecondGroup");

Utility Functions

There are two utility functions: the Gamma and Beta functions, which are used in statistics to calculate distribution values. These functions always return a double value and use one or two double values for input.

Example

This example demonstrates how to calculate the Beta function.

Visual Basic	Copy Code
Imports Dundas.Charting.WebControl ... ' Calculating the Beta Function. Dim beta As Double = Chart1.DataManipulator.Statistics.BetaFunction(0.5, n/2.0)

C#	Copy Code
using Dundas.Charting.WebControl; ... // Calculating the Beta Function. double beta = Chart1.DataManipulator.Statistics.BetaFunction(0.5, n/2.0);

See Also