Orange Data Mining with Spectroscopy
The CLS Mid-IR beamline, in collaboration with infrared beamlines at SOLEIL and Elettra, has developed open-source spectral data processing and analysis software tools building on the http://orange.biolab.si multivariate data analysis, visualization and mining suite.
Spectral processing routines such as baseline subtraction, normalization, FFT, peak integration, differentiation and smoothing can be quickly and easily applied to a dataset. The resulting spectral features can than be combined with other sample attributes such as concentration, growth conditions or other perturbations for multivariate analysis using tools such as regression and classification models, principal component analysis, hierarchical cluster analysis and more.
This software is developed in Python and is available under an open-source license, enabling easy collaboration, contribution of new features and verification of processing algorithms. Source code and development discussions for the Orange Spectroscopy add-on can be found at the github page.
The preferred installation method is to use the pre-built miniconda installer. This provides the 64-bit version to handle large image datasets in a user-friendly installation format.
- Download Orange from: https://orange.biolab.si/download
- Windows: Orange Downloads Windows choose
Orange3-3.x.0-Miniconda-x86_64.exe(the default, unless you have a 32-bit OS)
- MacOS: Orange Downloads MacOS
- Linux: Orange Downloads Linux
orange-spectroscopyby going to "Options" "Add-ons..." and choosing "Spectroscopy".
Note that if you already have Anaconda/miniconda installed, as of 3.9.0 the Orange installer will nicely re-use that installation.