Variants of a surname complicate reports that summarize family lines. But there are features to include variants (and statistical outliers) in with the larger grouping.
How do I group similar surnames?
I do not know how many users take advantage of Gramps' Group As surname override found in the Name Editor. It is probably a small minority.
But I have been configuring certain names to take advantage of the capability.
Accessing the Name Editor
Leveraging the Group As feature
For those not familiar with the function, surnames like “Clark” and “Clarke” can be grouped in the Grouped People view under a single grouping. I chose to bring the names together under “Clark”. With the Display name starting with the given name, this brings together every record regardless of the spelling of the name “Clark”, The names of “John Clark” are seen together with the names “John Clarke”.
In the reports that I use most when sharing with relatives (the Narrated and Dynamic Web), navigation by Surname or Individuals utilizes this grouping and all the Clark and Clarke individuals are grouped together. But the Surname heading was just the one name used in the grouping, “Clark”.
Then I had a revelation. There is no need to use just one or the other name to group together. So instead of just setting the override for the “Clarke” records to “Clark”, I have now started to set both overrides, Clark and Clarke to “Clark (Clarke)”. McDonald and MacDonald are now grouped as “McDonald (MacDonald)”, etc.
And these settings carry over into the two web reports.
Finding all the Group As Persons
Next, I needed to find all the Group As modifications that I had already made before this Aha! moment.
I did an un-compressed XML backup to explore for what Gramps was doing with the information. All globally changed Group As entries are stored at the end of the file in a nice compact list. This makes it easy to find the records that I had already done.