Visualizing statistical analysis of curve datasets in ParaView

A method for the visualization of datasets of curves has been implemented in ParaView. The method is based on performing a PCA (Principal Component Analysis) of the curves; then quantiles and outliers are estimated in the plane defined by the two more relevant components of the PCA; finally this information is reprojected in the original curves space. We have created 3 new Para View views, 3 filters and modified the existing PCA filter to also perform a robust PCA. Results of the analysis of curves coming from the nuclear industry are shown.