Tag Archives: modeler

Ranking algorithm for the layman – New Modeler Extension

There are a lot of ranking algorithms for you to use and represent your data. Now, that doesn’t mean that the ranking algorithm is the most appropriate for your situation. It all depends on what you are trying to accomplish and who will be the audience of your final product. Some ranking algorithms could be too sophisticated to be actually understood by a layman’s type of person. The reasoning to develop this particular piece of work was to not only rank specific type of datasets but be able to explain how the ranking is performed for the layman’s type of person.

main-view

The inspiration for creating this node came from this site -> http://www.psychstat.missouristate.edu/introbook/sbk14.htm. Now that this node exists, it allows us to reuse it over and over. According to the author of the site, transforming raw scores into percentile ranking, will enable us to: 1) Give us meaning and interpret the scores and 2) Provide a direct comparison between scores.

Sample Output:
results-after-ranking

If you want to learn more and try this extension go to the GitHub Repository. You can find my extension and all others in thePredictive Analytics Gallery.

Source: Missouri State – Score Transfromations

Source: | Ranking algorithm for the layman – New Modeler Extension

IBM SPSS Modeler Streams with Excel export nodes fail with Insufficient memory for JVM – United States

From time to time you might run into some issues with SPSS Modeler of running out of memory on Java Virtual Memory.  The way you fix it is by expanding the allocated JVM.

So, in order to do this we mustdo the following…

  1. Close SPSS Modeler if you have it open
  2. Open C:\Program Files\IBM\SPSS\Modeler\\config\jvm.cfg as Administrator
  3. Find the line with the following
    options, "-Xmx256m"
  4. Now change "-Xmx256m" to a new value where 256m = RAM.  So if you would like to lets say allocate 1G of RAM, then you will change it to options, "-Xmx1024m".

Source: IBM Streams with Excel export nodes fail with Insufficient memory for JVM – United States