November 2010

Hello Analytica Users!

Here is the November edition of the Analytica Newsletter, the place to stay up-to-date with Analytica tips and hints, new features and releases, training opportunities and modeling resources. This may be your last chance for a little light reading before the holiday rush sets in! Thanks for your support.

Thanks,
Paul Sanford
Analytica Newsletter Editor

Let your models see daylight with Analytica Web Player (AWP)

Does your local group of Analytica users sometimes feel as cozy as a cluster of cut off copper miners? Analytica Web Player lets you share your models with anyone with a web browser! And best of all, it’s free for Analytica users with active support for the Professional edition and above. Need more? You have the option of purchasing additional user session credits. Or, for larger scale applications, the Group Plan subscription provides additional functionality for collaborative projects and more user session credits.”

AWP Free Basic Plan

The Basic plan allows you to post an unlimited number of models and email URLs to anyone you would like to share the model with. Viewers can change inputs and generate results via the web. It includes 25 CPU-minutes per month.You can purchase additional credit blocks as needed ($25 for 25 CPU-minutes).

This plan is also offered as a 30-day free trial to all users regardless of support status. Sign up on the trial registration page.

AWP Group Plan: $200/month or $2000/year

In addition to basic plan offerings: Administrators can use multiple directories and control access to individual models with password protection. End users can browse models, change inputs, generate results, and save sessions for future reference. It includes up to 250 CPU-minutes per month. You can purchase additional credits at $100 for 200 CPU-minutes.

For more information see, Analytica Web Player.

Analytica 4.3 beta is here!

Analytica 4.3 beta is downloadable for the first time this month! Users with active support are invited to explore the latest version of Analytica and participate as beta testers. Two major focuses of this release are:

Domain Attribute Refinements

The Basic plan allows you to post an unlimited number of models and email URLs to anyone you would like to share the model with. Viewers can change inputs and generate results via the web. It includes 25 CPU-minutes per month. You can purchase additional credit blocks as needed ($25 for 25 CPU-minutes).

Structured Optimization

The Optimizer edition of 4.3 offers a more convenient and intuitive way to set up optimizations. You can now structure all types of optimization (LP, QP, NLP) using similar functions and diagram structure. Analytica 4.3 automatically figures out which optimization type is needed for your problem formulation. Most conveniently for complex models, it automatically combines all the decision variables (scalar and array) together to define the optimization. A new Object class, the Constraint node, allows you to specify constraints more efficiently.

See What’s New in Analytica 4.3: For download links and beta testing information see the Beta Tester Page on the Analytica Wiki:

Analytica Handy Hint: become master of your own density

You may have noticed that probability density graphs can sometimesappear choppy, especially if you have a small sample size. The built-in probability density function uses histograms– it sets up a set of bins and counts how many points land in each bin. A choppy appearance can sometimes be a natural outcome of this process. Choosing a larger sample size is one way to address the problem. Pull down the Result menu and select Uncertainty Options, Uncertainty Sample. Sample sizes of 1,000 to 10,000 are usually sufficient to produce smooth PDF curves. You can also tryusing more Samples per PDF step interval using the Probability Density analysis option. 100 to 300 samples per interval usually works well. If you want to smooth your PDF curves wi thout increasing sample size, you can use a simple algorithm known as a kernel density estimator. The algorithm also sets up a series of bins, but it blurs each point using a bell-shaped blurring function. The result is a smoother plot.

You will find a description of this technique in the Analytica Wiki, along with a downloadable demonstration file titled Kernel_Density_Estimation.ana.

In other news

Follow Us

Social networking is a new way to stay in touch with the Analytica community.Follow us on Facebook, Twitter and Linkedin.

Analytics Magazine: Nov/Dec issue now available

The latest issue of Analytics includes articles on Quantitative Consulting , rapidly adaptive decision making, and practical ways to drive customer field service. . Read more Header