Centurion

What is Centurion ?

Statgraphics Centurion is a comprehensive Windows desktop product for statistical analysis, data visualization, and predictive analytics. It contains more than 260 procedures covering a wide range of data analysis techniques. Statgraphics features an easy-to-use GUI that doesn't require learning a complicated command language. Especially useful for professionals is StatAdvisor, which explains the results of statistical analyses in a way suitable for presentation to non-statisticians.

Attribute capability analysis statlet

This Statlet performs a capability analysis using attribute data. The data can consist of the number of nonconforming items in a sample or the total number of nonconformances if an item can have more than one nonconformance. The analysis is based on binomial or Poisson distribution. Statlet will calculate estimates of parameters and confidence limits or upper confidence limits, capacity indices (best estimate and upper limit), and DPM (defects per million). The analysis can be based on a classical or Bayesian approach.

Big Data

To deal with big data, a special file type called statgraphics big data file was developed. These files have the .sgb extension instead of .sgd. They differ in two important aspects of statgraphics standard data files:

·          It stores numeric data in binary format instead of text. This avoids the step of converting each data value to a number when it is read in the program

·          Data is stored column by column instead of row by row. This drastically reduces the runtime when individual columns are read into memory.

Using SGB files, Statgraphics is able to analyze datasets that consist of many millions of records and thousands of columns.

Double-density statlet

The bivariate Statlet displays the estimated density function for 2 columns of numeric data. It does this using a two-dimensional contour plot or a three-dimensional frequency histogram. The joint distribution of the two variables can be considered multivariate normal or estimated using a nonparametric approach.

Butterfly Plot

The Butterfly Plot comparts two sample attribute data. It consists of two sets of bars that show the frequency distribution of each sample in a set of categories.

Capacity control charts

This procedure builds Phase II statistical process control charts to monitor capacity indices such as Cp and Cpk. Given a process that is considered capable of meeting declared requirements based on variable data analysis, these charts monitor continuous compliance with those requirements.

Capacity Control Chart Design Indicator

Statlet helps analysts determine the size of samples when constructing capacity control charts. Capacity control charts monitor processes that have proven stable and capable of producing results that generate a small number of non-conformities.

Classification and Regression Trees

The Classification and Regression Trees procedure implements a machine learning process to predict data observations. Creates models in 2 ways: classification models and regression models. Models are constructed by creating a tree, each node from which it corresponds to a binary decision. Given a particular observation, one travels through the branches of the tree until a terminal leaf is found. Each leaf in the tree is associated with a predicted class or value.

Definitive screening designs

A new type of experimental design has been added to the Experiment Design Wizard. Called Definitive Screening Projects, these projects are small projects capable of estimating models involving linear and quadratic effects, although second-order interactions are partially confused with each other and with quadratic effects. In addition, designs for 6 or more factors collapse into projects that can estimate the complete second-order model.

Demographic Map Viewer

This Statlet is designed to illustrate changes in location statistics over time. Given data in each of the local k stakes during p-time periods, the program generates a dynamic view that illustrates how the data has changed in each location. The data in each location is drawn using a bubble whose size is proportional to the value of the observed data. SHP files can also now be used to draw boundaries.

Distribution Adaptation for Arbitrarily Censored Data

This procedure analyzes data in which one or more observations are not accurately known. In particular, observations can be censored on the right, censored on the left, censored by interval or a combination of the three types. The procedure calculates summary statistics, adjusts distributions, creates graphs, and calculates a nonparametric estimate of the survival function.

Equivalence and Non-inferiority tests.

It is used to compare two independent means, compare two paired means, compare a single mean with a target value, and analyze the results of a 2×2 crossstudy. Unlike standard hypothesis tests that are designed to prove the superiority of one method over another, equivalence tests are designed to prove that two methods have essentially the same meaning.

Heat map

This procedure shows the distribution of a quantitative variable over all combinations of two categorical factors. If one of the two factors represents time, the evolution of the variable can be easily visualized using the map. A gradient color scale is used to represent values of the quantitative variable.

Monte Carlo Simulation - Tornado of Sensitivity

This graph shows the effect of each input variable on a response when it changes over a specified percentage of its probability distribution, with all other variables maintained in their median values. Variables are sorted from top to bottom in the order of their effect.

System Requirements

Windows XP, Vista, 7, 8, or later

At least 512MB of RAM

260 MB or more of available disk storage

CD-ROM drive or Internet access