June 15, 2015

Sample Viewer in pSeven

Updated on 20.11.2018

We continue to reveal new capabilities of the forthcoming pSeven 6.0 release. This time new and very important feature – Sample Viewer.

Sample Viewer gives quick view of different perspectives on a selected data sample (e.g. imported from CSV file or stored results of a workflow run), helps to gather insights from data.

Sample Viewer extends existing Data Table and Descriptive Statistics, with Correlations and Dependency tabs.

The Correlations tab is used to spot correlations in the data sample computing various measures of dependency including Pearson's, Spearman's, Kendal's and Distance correlations (with partial versions where available) and Mutual Information. For one to check if different types of dependencies exist between features.

Using the Dependency tab, one may quickly check how well the data sample is approximated with linear or quadratic model and assess quality with 'predicted vs. reference' scatter plot visualization and cross-validation error metrics. There's also information on features importance (column charts).