ProSensus, Inc.

Reveal the wealth of information contained within your data.

	Visualize your data through coloring options. Selections in score, SPE, Hotellings T2, and Observed vs.  Predicted plots are mirrored in all open plots from the same model. Easily interact with the data contained in your models. Exclude selected outlier observations with a mouse click.
Observe variable selections throughout your model. See how a variable or group of variables clustered within a loading plot are rated in a VIP plot. Remove selected unimportant variables with a mouse click. View raw data by making selections and clicking a single button.
	View model performance with R2 (%explained) and Q2 (predictive ability). Observe how well individual variables can be explained and predicted. Exclude variables that are poorly explained or have low predictive ability.
Simulate online monitoring of batch processes Create monitoring plots of various statistical quantities such as: SPE, Hotellings T2, Scores and Predictions of quality variables. Clicking the "Play" button displays the evolution of the batch.
	Easily handle unaligned data with the Alignment Tool. Simple addition/removal/modification of imported event tags. Define how the different phases in the batch can be aligned. Extract important batch features with landmark selections. *Included in ProSensus MultiVariate - Batch Edition

New in version 11.08:

• The ability to save an alignment done with the alignment tool. You can reload this alignment through Edit->Edit Alignment or when importing a batch block in the creation of a new project.

• Raw Data plots now have active plot options when the plot is an observation based plot with one variable

  plotted or

  a trajectory plot with one variable and one observation selected.
• There is now a new phases configuration dialog that appears when imported phases are not consistent that lets the user remove unwanted phases and reorder them.
• Model Properties dialog now displays all observations and whether they are included or excluded using the red and green column rather than showing only the ones included.
• When creating a contribution plot, if the percentage of observations that are missing for a variable are below a threshold set in the Properties menu, the contributions will be done with the X-Hat (imputed) values rather than placing a missing bar.
• Currently selected observations on plots are now automatically selected on newly created raw data plots.
• Currently selected observations or variables on plots are now automatically selected on new tables create from plots.
• Bug fixes and the Help has been updated with more information.

For more detail on some features, see the change log from within ProMV (Help->About) or see the ProMV help documentation.

New in version 10.08:

New in version 10.02:

• Import data from Microsoft Excel worksheets, CSV and other generic delimted text files.
  • *Excel 2007 is required to be installed in order to read Excel 2007 XLSX files.
• Export model parameters and results to Excel or Matlab with a mouse click.
• Subsets of data may be defined (for example: different recipes, different process conditions, normal and abnormal operation). Models may be built on one subset and applied to others.
• Batch trajectory data are unfolded in the most appropriate manner (see References) for batch modeling, monitoring and control (in a PCA / PLS modeling framework).
• Multiblock data structures are handled explicitly, which aids in understanding data on different aspects of your process. Typical data blocks include raw material properties, environmental conditions, snapshot and qualitative data, continuous process conditions, batch process trajectories and product quality data.
• ProSensus MultiVariate performs principal component analysis (PCA) on process data.
• ProSensus MultiVariate performs partial least squares (PLS) on your process data to relate them to your final quality properties.
• The use of PCA and PLS allows powerful techniques to be used to infer missing data in the training worksets and any prediction worksets.
• Detailed and transparent documentation on the methods used for calculations, with references to key l iterature.
• Tutorial documentation - takes you through the steps for mining your data.
• Enhanced plot interactions.
• And more...

Stay tuned for more feature announcements!