I have a DB with 70,000+ entries that’s been built over several years. In attempting to “housekeep” the file I recently discovered your tool that counts “Separate Trees”. I have more than 200 most of them with 1 entry, but a few with many and one tree with 250+.
Are these names duplicates with those in my main tree or separate entries that can be deleted without affecting the main tree. I don’t see any Record numbers for the separate tree entries. …
Although possible, they’re likely not duplicates (Command palette > Duplicate search and merge). They’re just not linked into that largest tree that you’re thinking they should belong to.
Turn on Record numbers Settings > General settings (Gear icon) > Number to display > RIN
Agree with @kbens0n --Just my opinion BUT any that are just one entry, you need to look at and see why you included them ( unless you downloaded them from Ancestry etc )— for example I have a whole bunch of small files for variations of Brentson ( 1-3 people) and same for Bunn-- suspect 3rd g-gramps Brentson had a 1st wife and kids before marrying my 3rd g-grandma Bunn ( who was an orphan at marriage)-- so I collected info on any variation of those names in the area…
The 250 tree-- definitely check that out as there must be some kind of connection-- my large tress are things like a guy who could actually be a son of my g-gramps by a 1st wife ( or definitely related somehow)–just haven’t found the proof…
A great deal of how to deal with separate trees in your RM database comes down to whether you only want to include people in your database for whom you have good evidence of a connection or whether you want to have more of a research database that includes unconnected people.
For example, if I research a cemetery for burials with the Bryan surname, I will enter all the people from the cemetery with the Bryan surname into my one and only database. Often, some of the newly entered Bryans will be in a separate tree within my database, and possibly within a tree containing only the one person. Then resolving if or how that person is connected to me becomes a to do task for further research.
Some users prefer to keep everybody in one database. Other users prefer to keep such unconnected people in a separate research database. I’m not sure which approach is best, but everybody needs to make a personal decision. If your decision is only to have one database, then I think you are bound to have these kinds of unconnected trees.
Good thoughts- I did it and removed lots of duplicates. Thanks.
I was hoping the one-entry trees issue would be resolved via the
I’m happy to begin agaiineliminating duplicates process. It had little or no effect - Damn!
Now the hard part remains - why do I have the bigger ones? I dunno. Can I assume none of these names are in my big, main file/database?
I’m happy to pitch them all provided I’m not deleting any names in the main DB.
It’s hard to know. Do you know how the people in any of the bigger disconnected trees got into your database in the first place? Perhaps you imported them from somewhere? Perhaps they used to be connected and perhaps you deleted someone in your database and in so doing the links were broken? You will just have to look at the people in the bigger disconnected trees to see how they might possibly be connected to you even though at the present time they are still in a disconnected tree.
The links that keep people in the same tree are parent-child links and spouse links. So for example, your first cousins are in the same tree as you by following these links: yourself your parents (parent-child link), the link from your parent your grandparents (parent-child link)m the link from your grandparents to the parent of your cousin (parent-child link), and the link from the parent of your cousin to your cousin. Some of the trees in RM databases can have much more convoluted linkages that involve spouse links, and and the links can be especially convoluted when cousins marry or when a person with multiple spouses marries a person with multiple spouses.,
Any Accidental or intentional change of a relationship…or removal of a person… can cause “TREES” — because there is not longer a path back to the home relationship person.
I almost have around 5 or so. I usually do not remove anyone just because I do not know why/how they are there. I only remove after I am highly confident they do not below and I will not find connection. As far as researching before deciding to remove – I have found that Family search can provide hints / clue of how you are related. (You will have need to put yourself/parents and your grandparents… usually from there it figures out where you belong in the global tree. If the person has a FamSearch ID (or can find on ) you can then “Show on Family” then click “View Relationship”
then it will show if and how you are related (in the global tree) like this
and also a relation chart of how you are connected ( cavaet) you need to verify each couple up and down to make sure they belong etc or you could add garbage
For sure do a backup before any operations that affect multiples of people. 
Thanks for the counsel… RM question: Is there a convenient way to determine if I can delete an entire peripheral tree without deleting anyone in my main tree? I’ve just removed all duplicates.
/Peter
No such functionality. Although RootsMagic counts the members of these separate trees, it does not currently have a concept of “main tree”. It doesn’t enforce limitations or narrowing because it’s a database capable of one or many trees.
Yes. Make a group out of the peripheral tree and then use RM9’s new tool to delete everybody in the group.
To make the group, highlight any person who is in the peripheral tree and use a marking criterion of Everyone in highlighted person’s tree.Having done so, it’s very important to look at the group to be very sure that it contains exactly the people you expect it to contain. If something has gone wrong with the marking process, then deleting the group could have disastrous consequences.
Two convenient ways to look at a group are to filter the left sidebar index by the group or to filter the mail People List by the group. If the group isn’t too large, I find either of the following as convenient ways to look at the group.
It doesn’t matter which person you highlight to initiate the marking process, but it’s usually convenient to highlight the person who RM lists as the representative of the tree in the Count Trees report.
Excellent…thanks guys… good ideas, suggestions, and cautions. I’ll look into this more tomorrow.
Another issue to check before you delete- do people in the separate trees have associations or shared facts with people in the main tree?
Thanks Richard - Another good point… /Peter
