What's New in CLUSTERSEER 2.0

The recently released ClusterSeer 2.0 includes a number of significant enhancements and new features and concepts. ClusterSeer 2 offers temporal methods and an expanded array of spatio-temporal and spatial methods. You can save project sessions, export images, import DBF files, import temporal data formats more easily, and load in other data to supplement maps. The most comprehensive cluster software available, ClusterSeer 2 contains 24 methods, including updated versions of all the methods previously found in Stat!.

New Features

Save Project Sessions

Save project session within ClusterSeer only. The project session includes the session log, and any corresponding maps and plots, as long as the plots were generated without using Monte Carlo Randomization techniques. ClusterSeer does not save the histograms or plots created from Monte Carlo runs. However, you can export them to a Bitmap or DBF format or run the analysis again to see the new histogram and plot.

Export Function

Export Images, Export plots, maps, and histograms to a Bitmap or DBF format. Once you have a Bitmap or DBF format, save it into any software program that accepts those formats, or any image processing program that can convert the graphics to different formats.

Load in Spatial Features

Spatial features are vector files that contain locations or spatial information but may not have associated data. Spatial features provide locations of natural or artificial boundaries or shapes to help visualize spatial data.

Restart Session

To restart your ClusterSeer session from scratch, choose "Restart session" from the "File" menu to clear the Session Log of the summaries of actions previously performed and any notes you have added.

Expanded Legend Pane for Maps

Most ClusterSeer maps are displayed in a three-pane window. The left-hand window lists the active layers in the map. The middle window contains the map itself, and the right-hand window is the map legend. The right panel identifies the symbols for active map layers and may be expanded to view the full legend names.

Shapefile Requirements and Shapefile Export

ClusterSeer sends an error message if data do not meet shapefile requirements. Prepare data with a GIS data editor so that it does not contain self-intersecting polygons. A polygon is called "self-intersecting" when two or more of its borders intersect anywhere except their endpoints. In a couple of weeks, a function will be added for exporting shapefiles. This function will be available as a downloadable update.  

Import DBF files for methods that take text files. For spatial and spatial-temporal methods, ClusterSeer will prompt you to select which columns in the data file hold the relevant info. You must include labels when importing DBF files. Modified CuSum is currently the only method that does not require a label. Moran's I, Local Moran, Oden's I (Pop) and Grimson's methods will not work with DBF files.

Enter years using two or four digits ('89, for 1989) and have a mix of both types in your data file. When you use two numbers for a year, ClusterSeer will assume that the preceding numbers were "19." Thus, you can use dates with two and four numbers in the same file as long as the dates other than those in the 1900s have four digits

New Concepts

Nearest neighbor relationships are part of methods such as Jacquez's k-Nearest Neighbor and Cuzick and Edwards' methods. These methods consider whether events neighbor each other in space or in space and time, respectively.

Methods use different ways to randomize. These include: drawing from a multinomial distribution, drawing from a Poisson distribution, conditional randomness, randomization among spatial locations by swapping labels, multiplication by a random number to alter distances between points, assuming cases are allocated with equal probability across the time cells, shuffling time distances or adjacencies and shuffling time of occurrence of cases across case locations.

This statistic evaluates the significance of multiple sets of Monte Carlo simulations in Jacquez's k-Nearest Neighbor Method and in Cuzick & Edwards' Method. It combines the P-values across the number of tests you specify (k). Similar to the Bonferroni and Simes combined P-values, this statistic gives an overall probability that accounts for multiple comparisons.

© Copyright 2004-2010 TerraSeer Inc.