* Forcing a match when merging two large gedcoms

[JoFitz]

I'm doing a file merge/compare between two large gedcoms (1000+ individuals in each). There is a lot of overlap between the two.
Some couples and their children have been very nicely matched by by the comparing process, but other couples are unmatched (even when it is barn door obvious they should be, same couple, same marriage date, same children, same in-laws) and I can't force them to match up. This is true even when I've got a couple matched up, but it won't match up the parents.
When I try to match them I get this message:
'Family match not allowed: Families can only be matched if the parents of the families are already matched or at least do not conflict (for example, a family with a missing mother can be matched with any family in which the father matches)'

What should I do? Should I just import one  into the other and then do an ordinary merge/compare between individuals and families (which although it will be tedious, at least I know that I won't have created two sets of parents for some families which will have to be weeded out) , or is there some clever way of forcing a match when FH really really doesn't want to let you?

Thanks
Jo

ID:5702

[RSellens]

Hi Jo,
Somebody will probably be able to explain this a bit more clearly than me, but basically the reason that it has not auto matched up is probably due to the 'parents' and the families not matching up.
If you find the parents in the 'individuals' list you may find that they have either not matched or have matched to the wrong people. So your first port of call is to correctly match the parents up as the individuals. Then you should be able to go the the relevant family records and match those up.

Richard

[JoFitz]

Thanks Richard.
Even though I could force some of the individuals in question to match, I could not force it to match the spouses, so it looked as though they were married to the same person twice. I still can't fathom out why apparently identical family units were unmergable.
In the end, my strategy was:
1. go down the list of suggested merges and remove the non-matching ones (lots!)
2. make a note of all the people who were now married to the same person twice and had a duplicate set of the same children by this person.
3. merge the files (now over 4000 individuals)
4. go into the families records page and merge the duplicate couples and their children at this stage. FH was quite happy to accept the merge here where it hadn't been at the file merge stage.
5. If you've missed merging any of these duplicates, you will soon notice when you look at the all relatives diagrams and there are lots of swirly coloured lines linking lots of duplicate people who are all related to each other!
Hope this helps someone else.
Jo

[Jane]

One recommendation from me is to make sure you work through the new file, before you merge and make sure at a minimum the names are identical, for instance Henry R Smith will normally not match Henry Robert Smith, but if you go back to the 'new' file and change the name the automatch will the work and will often ripple the whole branch.