Find clusters in your Microsoft® MapPoint® data points...
Do you need to find natural groups your MapPoint datasets?
For example, do you have your customers plotted as pushpins, and need to find natural groups for the best utilization of your salesforce? Or perhaps you need to identify "hotspots" in specific product sales?
If so, MPCluster will solve your problems. MPCluster identifies groups or clusters in your MapPoint data. MPCluster can then draw a boundary shape around each cluster and/or mark each cluster's center with a pushpin. MPCluster supports most of MapPoint's different dataset types, but is typically used with pushpin (point) datasets or shaded area datasets.
MPCluster lets you set the maximum number of clusters to find. Minimum and maximum size limits can also be set for the clusters. These constraints can also be defined using a data field.
Benefits include:
- Find clusters from thousands of pushpins.
- Can work with all pushpins or just a specific pushpin set.
- Find clusters within a number of size constraints.
- Cluster centers can be weighted using a dataset data field.
- Clusters can be drawn (annotated) as boundary shapes and/or central locations.
- Ability to manage existing cluster annotation from previous MPCluster runs.
- Cluster assignments and parameters can be written to a Microsoft Excel workbook.
Additional features, include:
- Annotation (boundary shapes) support all of MapPoint's colors and line thickness settings.
- Ability to limit the maximum cluster radius.
- Ability to set the minimum and/or maximum number of pushpins per cluster.
- Ability to set the minimum distance between adjacent cluster centers.
- Ability to constrain cluster sizes using a dataset data field.
- Ability to estimate the number of clusters present.
- Ability to delete the annotation from previous cluster calculation run(s).
- Ability to allocate unused data points to their nearest cluster.
- "Draw circle" function draws circle annotation around previous cluster centers.
- Walk-through tutorial.
- Multi-threaded for faster processing on modern computers.
- Supports all geographic editions of MapPoint.
- Works with the latest MapPoint 2011.
- Windows 7, Vista and XP are all supported.
Click here to download MPCluster. This is a free trial version that will last for 14 days, and must be registered if you wish to use it beyond this period. MPCluster licenses can be purchased via the Secure Web Form for US$100 each. Other purchase options and volume discounts are also available.
Please use our contact form for MPCluster support.
"We tested the combination of MPCluster, Mappoint 2010, and Excel 2003
as a low-cost, cluster analysis solution with a small learning curve.
Our task was to analyze a sample of roughly 1000 German Biotechnology Firms and
sort them into clusters; we defined clusters as conglomerates of 5+ firms within a maximum radius of
50 miles.
MPCluster, the commercial geographic clustering plugin for Microsoft Mappoint utilizes the K-means algorythm
to find clusters of datasets imported into Mappoint. The cluster boundries, once calculated, can be exported to Excel.
After several runs, MPCluster found roughly 30 clusters, most of which made sense.
We had to use common sense to remove statistical outlyers and group or ungroup certain clusters,
but overall MPCluster did a great job of getting results quickly.
While MPCluster and Mappoint together do not replace a professional GIS, they can be used for basic
cluster analysis to get results quickly and at a low price point.
Also, in our experience, the development team was extremely responsive and professional.
We look forward to upcoming versions of MPCluster and recommend anyone interested in cluster analysis to
try MPCluster."
Andrew Isaak, Institut for Small Business Research, Mannheim Business School, Germany
