Skip to content

Dimensionality

The Dimensionality widget offers multiple options for the generation of high-dimensional analysis plots considering multiple variables of interest.

Figure: Dimensionality widgets

Variables, analysis mode and split mode for high-dimensional analysis can be selected from the control panel in the Dimensionality widget:

Figure: Dimensionality settings

In the Variables list of the control panel, at least two variables must be selected for matrix plots and at least three variables for PCA or tSNE.

The Analysis modes available are generation of matrix plots, principal component analysis (PCA) plots, and t-distributed stochastic neighbor embedding (t-SNE). These dimensionality plots are scatter plots showing single animal or mean data (according to split mode) per time bin (each dot represents one time bin) depending on time binning settings.

Different Split Modes (Total, By Animal, By Run, By Factor) can be applied to all analysis.

  • Split mode ‘Total’ uses the mean data values of all animals per time bin.
  • Split mode ‘By Animal’ does not calculate means but uses individual data points for each animal per time bin.
  • Split mode ‘By Run’ uses the mean data values of each run per time bin.
  • Split mode ‘By Factor’ uses the mean data value of each group of the selected factor per time bin.

Figure: Split mode and Update

Clicking Update in the control panel is necessary to display analysis results or to apply changes in the analysis settings.

Warning

Only animals selected in the Animal list are considered for the calculation and display of plots in the Dimensionality widget. Changes regarding the selection of animals are only applied after clicking Update.

The plot appearance can be edited by using the tools available in the plot menu at the bottom of the control panel.

Figure: Plot menu

From left to right: - Home button (‘House’): Reset the orientation of the plot in its frame. - Undo/Redo (‘Arrows’ left/right): Undo and redo actions regarding the orientation of the graph in its frame (moving and zooming). - Pan (‘Arrow cross’): Grab and move the graph within the frame of the plot. - Zoom (‘Magnifying glass’): Select an area of the plot to zoom in/ enlarge the selected area. - Subplots (‘Slider bars’): Adjust the dimensions of frame borders and spacings between graphs (in the case of multiple plots). - Customize (‘Graph’): - Define title, axis range, label and scale, and legend. - Adjust style, size and coloring of data points. - Save (‘Memory disc’): Save image on the hard drive.

Note

Any changes in graph appearance made through the plot menu are applied immediately, without clicking Update.

Figure: Dimensionality widget with control panel

Dimensionality plots can be exported via the Save button (‘Memory disk’ symbol) in the plot menu at the bottom of the control panel or added to the report by clicking Add to Report.