Partial least squares discriminant analysis (PLS-DA) is an adaptation of PLS regression methods to the problem of supervised1 clustering. It has seen extensive use in the analysis of multivariate datasets, such as that derived from NMR-based metabolomics.
Pathomx is a workflow-based data analysis tool built on IPython. It began as a metabolomic-analysis toolkit, but has extended to support general data analysis workflows. It aims to be simple to use for non-experts while powerful enough for complex analysis tasks. Key to both of these goals is the ability to create 'custom tools' that can be drag-dropped together to form larger workflows.
This notebook contains snippets of code that are useful when working with MATLAB in IPython Notebooks.
Now we've got the classes for the data plotted, we can now plot the mean values (of the spectra). To get a style assignment for a given class we need to call
.get_style_for_class() on the styles manager. This returns a useable style object that can output the keyword arguments needed to correctly style matplotlib plots.