MyHeritage import confusing people with the same name

Bugs Bugs: closed
#1

Does the MyHeritage segment data import use the sort of relatively new ID field to identify the match fully or just the name, which is not necessarily unique?
I’ve just imported the MyHeritage match and segmnt data for my own MyHeritage kit as a new DNA profile into my LIHClintonParkerAutin tree/project.
I have two matches of the same name, one managed by herself, the other by someone else.
The managed kit only shares 2 segments, largest 8cMs, the former significantly more and her father even more.
In match list view:

  • the self managed kit is showing no segment data and the correct match and tree links.
  • the managed kit tree and match links go to the correct managed kit’s tree and match but the segment data looks to include both people’s data.

In segment view
It looks to me that I’m getting segment data for both of the matches combined but the managed kit’s name, match and tree links.

#2

That’s correct unfortunately. The segment file export from MyHeritage matches on names only. I didn’t realize MyHeritage had added an ID field recently. I will change the code to use the new ID field for MyHeritage segment imports.

In the meantime, if you use DNA Gedcom to import MyHeritage segment information instead of importing the MyHeritage segment file directly, the MyHeritage segment file generated by DNA Gedcom includes kit IDs, so two people with the same name won’t be mixed up.

#3

Thanks Dallan.
Of academic interest now that the ID field is avaliable but in my own personal database of matches/segments before it was available I concatentated the match name and the managers name before matching.
Didn’t help where the duplicate names weren’t managed by others though.

Hope you’ve been training up FindMyPast staff well.
All the best in whatever exciting thing you’ll dream up next,

Lorna

#4

How did you get the ID field in the matching segments CSV file from MyHeritage? I just exported the matching segments file and all I got in the file were these columns:

DNA Match ID
Name
Match name
Chromosome
Start Location
End Location
Start RSID
End RSID
Centimorgans
SNPs

The “DNA Match ID” column has the ID of the profile (which is the same for every line in the CSV file), but the “Match name” column just has the name of the matching person - no ID. What did you do to end up with a matching segments file that has an ID for the matching person?

#5

My segment match data Match ID is not the same for every line, but differs for each match, so if your first match has a lot of segments, yes the first few rows will look identical, but the last 38 characters change for the next match.
The first 38 characters are identical for each row in the file

For my own database I split the Match ID column into two
Left 38 and Right 38 which gives you the person the kit belongs to (the left bit) and the match the segment data belongs to (the right bit)

#6

You’re right - I’d forgotten that’s how MyHeritage exports their IDs. I’ve made the fix. You’ll need to re-import your segment matches. Would you please let me know my change doesn’t fix the problem?

#7

confirm fixed Dallan thank you.
But you might like to update the import blurb on the segment upload page as a result - the least liked part of a developer’s job :slight_smile:

#8

Oops. Thanks for the reminder! I removed the warning about non-unique names from the MyHeritage segment upload page.