New Text Report for user testing. This report lists all of the DNA shared matches of the active person with the relationship and common ancestors. To use:
For each person with a shared match (as measured in cM - centimorgans) create an Association of type ‘cM’. Add a Note to the Association with the value (in cM) for the match. Optionally add a Citation to specify testing vendor (like Ancestry). This Citation is not used within the report.
Optional: identify where the match information was calculated. For each Associated Person as appropriate,
create a person Attribute ‘Ancestry ID’ with the Ancestry user name.
create a person Attribute ‘DNAkit’ with the GEDmatch kit number.
These Attributes are included in the report if provided. An associated person can have none, one, or both attributes.
Run the report from the menu: Reports → Text Reports → DNA Relationship Report. There are no special report options (at this point). If the output format is Plain Text, I suggest a characters per line of 200.
Is it not possible to leverage the notes created for the DNA Segment Map, and sum the cM within them? Or at least leverage those associations by attaching the new notes there instead of to new associations?
Is it possible to use it as a Quick View or is that something else entirely?
I do not understand the added utility of this feature. Isn’t the cM something that can be calculated from your existing DNA Association note too? Couldn’t it start there? Extract the cM data and compile a list.
I do have a big CSV file of matches downloaded from the FamilyTreeDNA family finder. And I’d love a way to ‘work’ that file. (With some features for doing row-based diffs against fresh downloads. No sense reworking the same people on each download.)
Fair question and I should have explained more. If you only did an Ancestry test, you do not have DNA segment info (and no DNA Association). You only have an overall cM value. In my case, I have over 130 connected people who have done an Ancestry test (and are identified in my tree) and only 25 of those who have also transferred their data to either FTDNA or GEDmatch to get segment info. Plus a small number of people who did a Family Finder test but not an Ancestry test (or a 23andme test and transferred to GEDmatch).
Brian - for the FamilyTreeDNA download you mentioned, that is the overall cM with no segment info. This is the data which I put into the cM Association Note. The challenge (as pointed out in another thread) is there is no mapping the person to the gramps ID. I take the Shared DNA field from this list and update the Note with this value (one-time) manually. I am open to other ideas how to automate this. I dont find that this list changes much over time. If I do get a new person, it is pretty easy to manually add.
In the case of geneanet, there is an ID that could be tied to a Person attribute (as mentioned in the other thread). That persistent ID would need to be a gramps Person Attrribute, I think. In my case, I cannot identify any of the geneanet connections to my tree (since geneanet is more popular in Europe, where I have less known connections).
For the people that have segment info, you could add the segment cMs to get the overall cM. But that is a small fraction of the people who have Ancestry data (the most popular test in the US).
And the other issue is that the shared cM value calulated by each vendor is slightly different (based on different rules), So an Ancestry calculated cM and a sum of the FTDNA segments will not be the same value. Sometimes the difference is large (more than 10%). I decided the principal of least astonishment would be to only read the cM Association and not calculate based on the DNA Association.
The other challenge I have is that I am unable to identify most of the FTDNA or GEDmatch matches to my tree (since the userbase for these generally do not include a tree, to help connect). In the case of Ancestry, a large fraction of the users include a tree, that can assist in locating the person in my tree. If I look at the percentage of Ancestry users at 20 cM or more, I have identified 90+%. The same measure for FTDNA or GEDmatch is less than 20%.
Makes sense, thanks for explaining. But I guess I would still prefer to have only one “DNA” Association per match, with one or more notes (total cM or segment-level cM) rather than two associations between the same two people.
Another thing you might consider at some point is summarizing known matches by their relationship distance and total cM, as in this chart on the ISOGG wiki (it appears on this page). Its Y-axis is based on the number of segments, but an alternative version could use the relationship distance as the Y-axis rather than as the color of the points (it would look different in that case).
Diving in to this, how would you do an associated person who has both an FTDNA and GEDmatch (or geneanet or …) list of shared matches? Is this a 2nd association? Or a single association?
If you do a single Association, then you would need to do 2 Citations with a Note attached to each Citation, rather than two Notes on the Association. And the view would be the OR of the two Notes, with no ability to separate the testing orgs. Is that getting too confusing?
If you do 2 Associations, you can see each in the combined and single chromosome view.
Personally, I was thinking this is two Associations but I am open to other opinions.
I think the cM Association vs also using a DNA Association needs to be consistent with the 2 test case.
From a logical perspective, I have one association with the other person – they are one of my DNA matches. And I have two different sources for that. Here I am using the terms “association” and “source” in their generic sense, not necessarily Gramps-specific, but that is what they would lead me to do.
Why would the view not be able to separate the testing orgs (perhaps not as it is currently coded)? Maybe I need to spend some time thinking about how I would want it to look in that case.
If it needs to be two associations for whatever reason, then I would just suggest that “cM” is not a very meaningful name for an association. Maybe something like “DNA match (summary)” vs. “DNA match (detail)”.
For a single Association, the legend would have one entry and one swatch color. So the two Note contents would not be color-differentiated.
One other point, if we use the idea that a cM Association is not needed for people with a DNA Association, and the overall cM will be summed from the individual segments, then the sum would be done over both Notes? Or averaged over the Notes? Or …
That would be OK. Ideally the hover text for a particular segment could indicate the source, But since the segments would be overlapping anyway, that doesn’t really matter. I guess it would be a problem to know which note to open when the user right-clicks on a segment.
They should probably never be summed. Instead of averaging them, could the text report just have two line items for that person, instead of one?
Having said all that, maybe my feedback is not that relevant. Since I don’t have DNA on Ancestry, I don’t have any matches for whom I know only a total amount. I use FamilyTreeDNA and GEDmatch, and for purposes of entering segment data in Gramps I would be OK with choosing one or the other as the source for a match who is on both,
Updated Report Gramplet to use DNA Association (as suggested). Note can now contain segment info (like DNAsegment gramplet uses) or a single number which is the overall cM (for an Ancestry test which does not provide segment-level info, for instance).
DNAsegment gramplet will ignore if the note contains a single overall cM value.
