20th Mar, 2012

New Releases: PLS_Toolbox and Solo 6.7, MIA_Toolbox 2.7

Updates to our flagship PLS_Toolbox and Solo were released last week; they are now in version 6.7. This is in keeping with our policy, (began in 2008), to release significant upgrades twice yearly. Our Multivariate Image Analysis (MIA) tools were also updated with the release of Solo+MIA 6.7 and MIA_Toolbox 2.7.

As the Version 6.7 Release Notes show, the number of additions, improvements and refinements is (once again!) rather long. My favorite new features are the Drag and Drop import of data files, Confusion Table including cross-validation results for classification problems, and Custom Color-By values for plotting.

PLS_Toolbox/Solo can import a wide variety of file types, and the list continues to grow. Drag and Drop importing allows users to drag their data files directly to the Browse or Analysis windows. They will be loaded and ready for analysis. For instance, users can drag a number of .spc files directly into Analysis. Forget some files or have additional files in a different directory? Just drag them in and they will be augmented onto the existing data.

The Confusion Table feature creates several tables summarizing the classification performance of models. This includes a “confusion matrix” giving fractions of true positive, false positive, true negative, and false negative samples and a confusion table which gives number of samples the actual and predicted classes. Tables are calculated for both the full fitted model and for the cross-validation results. The tables can be easily copy and pasted, saved to file, or can be included in the Report Writer output as html, MS Word or PowerPoint files.

With Custom Color-By users can color points in scores and loadings plots using any currently loaded data or with new data loaded from the workspace. For instance, samples in a PLS LV-2 versus LV-1 scores plot can be colored by the scores on another LV, their actual or predicted y values, leverage, Q residual, specific X-variable, additional Y-variable, or any custom variable from the work space. The allows deeper investigation into the cause of specific variations seen in the data.

Want to find out more about our latest releases? Create an account in our system and you’ll be able to download free 30-day demos. Want prices? No need to sit through a webinar! Just check our price list page, which includes all our products. Just click Academic or Industrial.

As always, users with current Maintenance Agreements can download the new versions from their accounts.

Questions? I’d be happy to answer them or refer you to our development team. Just email me!

BMW

Leave a response

You must be logged in to post a comment.

Categories