Using Functional Data Explorer for Spectral Data

Since the Functional Data Explorer was introduced in JMP Pro 14, it has become a must-have tool to summarize and gain insights from shape features in sensor data. With the release of JMP Pro 17, we have added new tools that make working with spectral data easier. In particular, the new Wavelets model is a fast alternative to existing models in FDE for spectral data. 

See how to:

• Understand what is different about handling functional and spectral data – measurements take over time
  • Sensor data (for example, fermentation)
  • Spectral data (for example NIR Spectroscopy)
  • Time Series data (for example, Consumer Price Index)
• Understand challenges to pre-processing Spectral Data
  • Raw data are not always clean for direct use
  • Typical problems are baseline, scattering, and noise
  • Standard Normal Variate,  Multiplicative Scatter Correction,  Savitky-Golay Filter and Baseline Correction pre-processing algorithms are available
  • Typical problems are baseline, scattering, and noise
• Understand data models available in JMP Pro
  • B-Spline -  A piecewise polynomial curve, and the knots are the points where the pieces meet for non-periodic data
  • P-Splines Basis – A piecewise polynomial penalized B-Spline  and the knots are the points where the pieces meet
  • Fourier-  For periodic (oscillatory) data
  • Wavelets (Haar, Biorthogonal, Coiflet, Daubechies and Symlet) –  For situations where the signal contains discontinuities and sharp spikes
• Analyze a Mass Spectrometry, Prostate Cancer case study to characterize changes in protein expression
• Analyze a HPLC Sophorolipids biosurfactants case study to optimize conditions for peak separation

Video 1: 

Video 2: 

Video 3: 

Resources

• Tutorial and demo on FDE by Chris Gotwalt, JMP Director of Statistical Development 
• Mastering JMP video and materials on Building Predictive Models for Spectral Data
• Mastering JMP video and materials on Understanding and Modeling Response Curves