Was just tweaking the Quilt Chart page on the wiki (because I’ve always found it a bit difficult to read) but it really didn’t have an introduction. So I visited the Aviz (Analysis and VIsualiZation) website that had the foundation documents.
And ran across their Bertin Matrix project that looked cool. It simplified a table of numbers by representing them with scaled dots.
It was a costly manual physical diagramming system in the 1970s & 80s… before spreadsheet software. It looked well-suited to adapting to a digital tool.
You could assemble rows with blocks that corresponded to a percentage match to the question in a column header.
The matrix diagram was locked if both column and row skewers were inserted. If the column skewers were pulled, then the row skewers could be easily re-ordered. Subsequently, if the column skewers were re-inserted and the row skewers removed, then the column skewers re-ordered.
The example below was a simple boolean “Oui ou Non” (in English: Yes or No; 100% or 0%) using Black or White blocks.
In Gramps, the Rows and Columns could be simulated with sets of custom rules. The Rows could return a subset of objects (like the Persons who are a 5th great-parent of the Active Person) and the columns could search those people for which have a certain attribute (like the person’s birth-fallback having a certain place enclosure). The percentage could be the count of Attribute matches divided by the count of Row objects. The scaling could be done by applying the percentage to the fontsize of a horizontally & vertically centered black circle character () where that 100% size fills the table cell completely.