MODDE Pro 12 includes state-of-the-art functionality for creating designs and analyzing the results with confidence, supporting the Quality by Design initiative through a set of useful tools.
• One-Click feature, including automatic outlier detection, transformation and model tuning.
• Generalized subset designs – A unique design setup that generates a sequence of reduced design sets.
• Improved advisor in the Analysis wizard.
• Definitive screening designs for 4 to 30 factors.
• Export function to SIMCA.
• Response correlation effect optionally included in Design Space calculations.
• Qualitative factors with missing level supported.
• Robust optimization • Probability estimates for a robust solution
• Design Space documentation
• Advanced and interactive setpoint analysis with capability estimation
• Proven acceptable range documentation
• Generalized subset designs for optimal complementary design sets
• Special application designs for example RED-MUP, Stability testing and reduced combinatorial
• Advanced constraints of design region
• Mixture designs
Detailed information about functionality in MODDE is available at our website: http://mksdataanalytics.com/product/modde-pro
1. It has now one-click analysis. Using One-Click in the Analysis Wizard aims at making the analysis process as simple for you as possible. Auto transform and Auto tune are performed without your involvement while there are other warnings that need your attention. This section explains a One-Click scenario where no interaction is necessary.
2. MODDE V12 is very easy in usage. MODDE is known for its graphical interface and interactive graphics. MODDE V12 has further made it easy to set up designs, enter data, and analyse the results.
3. New designs: When quantitative and qualitative factors are specified to have two levels, the design generation is straightforward. There are then several classical design families to choose from, for instance full factorial designs, fractional factorial designs, and Plackett-Burman designs. However, when quantitative and qualitative factors at three, four, or even more levels are involved in a DOE investigation, a challenge arises in the sense that the number of factor combinations grows substantially, far beyond what is required by e.g. a two-level fractional factorial design. For practical reasons, there may then be a need to limit the number experiments required. The generalized subset designs, GSD, is a new entry in MODDE providing a possibility to accomplish reduced designs even when handling multiple multilevel factors. This design setup generates a series of reduced designs, subsets, that are logically linked, such that, when combined, all subsets will add up to a full multilevel multifactorial design where all factor combinations are encoded by the global design. Conceptually, the output of GSD is similar to how two-level fractional factorial designs represent complementary reductions of two-level full factorial designs.
There is a much more detailed treatment of the statistics in this publication (submitted to Analytical Chemistry): Generalized fractional factorial designs - an approach to balanced subset selections in multi-choice analytical chemistry applications, written by Izabella Surowiec, Ludvig Vikström, Gustaf Hector, Erik Johansson, Conny Vikström, and Johan Trygg
4. New, design from scores: This is a typical QSAR application where a subset of molecules is selected by design. Molecules described by many variables are compressed to score vectors in SIMCA. These score vectors can then be imported directly from the SIMCA usp file.
5. MODDE V12 has excellent features in selecting less experiments with enough information for RMS, finding an optimum using quadratic terms. http://umetrics.com/services/upcoming-webinars/recording-available-looking-beyond-central-composite-designs-what-other and : http://umetrics.com/services/upcoming-webinars/recording-available-robust-optimization-part-2-11
6. Fit Models: The data collected by the experimental design are used to estimate the coefficients of the model. The model represents the relationship between the response Y and the factors X1, X2, etc. MODDE uses multiple linear regression (MLR) or Partial Least Squares (PLS) to estimate the coefficients of the terms in the model. MODDE recommends PLS when the investigation has a high condition number.
Do you wish to try MODDE Pro V12, please visit : http://umetrics.com/product/modde-pro “TRY NOW” at the bottom.
If you wish to have training, please visit : http://umetrics.com/services/courses Two courses can be of interest:
- Design of Experiments, 2 days 2017-05-16 Malmö, Sweden
- Quality by Design and Design Space, 3 days 2017-05-30 Basel, Switzerland
Do you have any questions please contact Jan Smit at firstname.lastname@example.org
Laat uw naam en telefoonnummer achter en wij bellen u zo spoedig mogelijk terug.