Data Events
- The event count in the selection details gives the user a quick idea about the selected data:
- How many mutations (categorical or binary data) ?
- How many over and under-expressed values (continuous data)?
- How many significant p-values (test results)?
- How many siginificant z-scores (z-score enrichment results)?
When selecting a rows and columns, Gitools shows some basic aggregations and counts the data events within the
selected data. Data events are a way to binarize the data on the fly which is also used f.ex. by the Mutual exclusion and co-occurrence test.
Which data points are considered as events and which aren’t depends on the color scale that is being used to show the data.
Hovering the mouse over the Events drop-down selector in the selection details box will give you more information
and lets the user change what data points are considered events or not.
Examples:
- Categorical scale: All values represented in the categorical scale are events. The rest of the values are non-events.
- P-value scale: All values falling below the significance threshold are events. Non-significant and null values are non-events
- Linear scale: All values falling below minimum and above maximum of the scale are events.