Showing relationship pathways between individuals

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Showing relationship pathways between individuals

brian fitzgerald
I am not having any luck getting  a relationship pathway displayed using rel_graph when the relationship is not exclusively descendant on both pathways. Better said, it seems to work OK when there is a direct common ancestor (ie one common pivot node person), but if one of the individuals need to descend a couple of generations to arrive where there is a common ancestor present (ie a second pivot node in the pathway) (the pivot node person is usually the result of  a marriage) it fails.
The filters I am using  are the "pathway between an individual and the result of a filter" and the result of the filter is obtained by the simple "descendant of an individual". I thought this was going to achieve my goal but it crashed my Gramps. I filed a bug report but I want to know if I am asking Gramps to do something not designed in. Perhaps an edge case which others don't need.

If someone has a  recipe using "filter math magic" which they would share I would be grateful, as might others.

Additionally, I suppose that two people connected by a common descendant have a 'relationship' as well as those who share an ancestor. I am not entirely sure what you would call them, however. This aspect may be useful moving forward as DNA admixtures need to be tracked and triangulated, so Its not so much of an edge case in reality.

Thanks
 Brian

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Showing relationship pathways between individuals

Doug-11
On 20/11/17 17:56, brian fitzgerald wrote:

> I am not having any luck getting  a relationship pathway
> displayed using rel_graph when the relationship is not
> exclusively descendant on both pathways. Better said, it
> seems to work OK when there is a direct common ancestor
> (ie one common pivot node person), but if one of the
> individuals need to descend a couple of generations to
> arrive where there is a common ancestor present (ie a
> second pivot node in the pathway) (the pivot node person
> is usually the result of  a marriage) it fails.
> The filters I am using  are the "pathway between an
> individual and the result of a filter" and the result of
> the filter is obtained by the simple "descendant of an
> individual". I thought this was going to achieve my goal
> but it crashed my Gramps. I filed a bug report but I want
> to know if I am asking Gramps to do something not designed
> in. Perhaps an edge case which others don't need.
>
> If someone has a  recipe using "filter math magic" which
> they would share I would be grateful, as might others.
>
> Additionally, I suppose that two people connected by a
> common descendant have a 'relationship' as well as those
> who share an ancestor. I am not entirely sure what you
> would call them, however. This aspect may be useful moving
> forward as DNA admixtures need to be tracked and
> triangulated, so Its not so much of an edge case in reality.
>
> Thanks
>  Brian
Can you give a couple of examples (anonymous) and say
exactly what means you're using to display a relationship?
Using Deep Connections I can't find any problem however
indirect the connection.

Doug

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Showing relationship pathways between individuals

Doug-11
Thanks for that clarification.

Re your P.S.: I'd like to see your draft, as it may have
something in common with an old feature request for a
matrilineal filter.

Happy Thanksgiving to you!

Doug


On 22/11/17 19:19, brian fitzgerald wrote:

> I reported a bug 53133 and Paul Cully wgangled it. See below
> exerpt from the resolution.
> .
>
> .
> A NOTE has been added to this issue.
>
> ----------------------------------------------------------------------
>   (0053133) prculley (developer) - 21-Nov-2017 17:28
>   https://gramps-project.org/bugs/view.php?id=10279#c53133 
> ----------------------------------------------------------------------
> Thank you for the test tree.  It is quite a convoluted tree, and stressed out
> the deep relationship filter considerably more than the original author
> expected.
> I rewrote the filter to use a different algorithm and had it print out the
> longest 'path' it explored, turns out it was 675 persons long.  The original
> algorithm broke at around 500 long.  A 'path' is from person to person through a
> family, looking for the connection between people.
>
> If you want to experiment with the fixed filter, let me know and I will tell you
> how to get and patch it in, tomorrow.  Otherwise you will have to wait for a
> newer version of Gramps.
>
> .
> .
>
> The posting to the forum was a bit more generic but related in the sense that I was not sure that the algorithm was trying to
> look both up toward a common ancestor and/also down toward a common descendant. Also there are two interesting edge cases in this algorithm
>
> 1) can the algorithm look for a pathway where there is both an up pivot individual (a common ancestor) and then back
> down the other leg of the  relationship to a second pivot individual and thence back up again. That is a lot of potential
> pathways to cover in a even a moderate size tree?
>
> 2) is a marriage without issue a valid relationship (think pivot point individual!) where the pathway can cross over? and if
> not can/should it be selectable as an option.
>
> In any case my Gramps was crashing at the first ultra-long pathway it found in my database(675 persons long) before it
> found some shorter more useful pathway so I never could test those edge cases.
>
> It also occured to me that I didn't know the name of an "inverse cousin", that is someone on
> the upside pathway to the common descendent. Idle curiosity really.
>
> I can wait for the next Gramps as Paul puts it. Should be long now until 5.0beta1, I hope.
>
>
> Thanks.
>
> Brian
>
>
> PS I have a draft posting for the forum which I am hesitating to send since it deals with
>
> Data and ontology structures which might require some more programmer/designer specific
> knowledge. It is a possible 'feature request' related to genealogical features found in tribes and
> kindreds. Probably an under-served user community.
>
> If I send it to you would you 'vet' it before I bother everyone. Else you can tear it apart
> in the Mantis Feature requests over the weekend. Happy Thanksgiving!
>
>
> Brian
>
>
> On Wed, Nov 22, 2017 at 11:41 AM, Doug
> <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     On 20/11/17 17:56, brian fitzgerald wrote:
>
>         I am not having any luck getting  a relationship
>         pathway displayed using rel_graph when the
>         relationship is not exclusively descendant on both
>         pathways. Better said, it seems to work OK when
>         there is a direct common ancestor (ie one common
>         pivot node person), but if one of the individuals
>         need to descend a couple of generations to arrive
>         where there is a common ancestor present (ie a
>         second pivot node in the pathway) (the pivot node
>         person is usually the result of  a marriage) it
>         fails.
>         The filters I am using  are the "pathway between
>         an individual and the result of a filter" and the
>         result of the filter is obtained by the simple
>         "descendant of an individual". I thought this was
>         going to achieve my goal but it crashed my Gramps. I
>         filed a bug report but I want to know if I am asking
>         Gramps to do something not designed in. Perhaps an
>         edge case which others don't need.
>
>         If someone has a  recipe using "filter math
>         magic" which they would share I would be grateful,
>         as might others.
>
>         Additionally, I suppose that two people connected by
>         a common descendant have a 'relationship' as well as
>         those who share an ancestor. I am not entirely sure
>         what you would call them, however. This aspect may
>         be useful moving forward as DNA admixtures need to
>         be tracked and triangulated, so Its not so much of
>         an edge case in reality.
>
>         Thanks
>          Brian
>
>     Can you give a couple of examples (anonymous) and say
>     exactly what means you're using to display a relationship?
>     Using Deep Connections I can't find any problem however
>     indirect the connection.
>
>     Doug
>
>


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Showing relationship pathways between individuals

prculley
The posting to the forum was a bit more generic but related in the sense that I was not sure that the algorithm was trying to
look both up toward a common ancestor and/also down toward a common descendant. Also there are two interesting edge cases in this algorithm

1) can the algorithm look for a pathway where there is both an up pivot individual (a common ancestor) and then back
down the other leg of the  relationship to a second pivot individual and thence back up again. That is a lot of potential
pathways to cover in a even a moderate size tree?

The algorithm looks in all directions, up to parent families, both parents and siblings, and down to a persons families.  It does NOT look through 'Associations', those people on the 'Associations' tab of the person editor.

2) is a marriage without issue a valid relationship (think pivot point individual!) where the pathway can cross over? and if
not can/should it be selectable as an option.

The algorithm doesn't care if a family has no children (or spouses, or parents) it still goes through the available linkages.  So a spouse in a childless family is a valid linkage.  There are no options to change this.

Paul C.


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Showing relationship pathways between individuals

Doug-11
In reply to this post by Doug-11
Brian,

I see that what you're contemplating is a lot more complex
than I had in mind.
I think you should indeed put it up for discussion, but I
suspect there's meat there for a whole set of feature requests.
Good luck!

Doug



On 25/11/17 18:30, brian wrote:

> Doug ,
>
> the text is pasted below. It is wordy because I think there
> should  be a rationale for any feature requests.
>
>> I am mostly working in the area of medieval Irish
>> genealogy and my own intersection with it. I am preparing
>> for an onslaught of genetic DNA test results and newly
>> translated genealogies in the next few years which will
>> link modern Irish descent people back to their very
>> ancient roots.
>> These aspects are very high interest right now.
>> There are several problem domains in Irish Genealogy and
>> Gramps does seem to be addressing them head on...dated
>> place names, Patronymics, Family structures, events,
>> diacritics, extended family structures, etc. but there is
>> one other emerging recently which could be addressed,
>> fairly easily in my opinion, which would assist many other
>> ethno-genealogical users. Tribes and kindreds.
>>
>> Ireland is not alone having this this rich historic tribal
>> heritage.
>> This, by the way, closely tracks genetic genealogy, a fast
>> moving field.
>>
>> I would like to have the option to describe the clan
>> hierarchies for an individual in Gramps and wonder whether
>> there is a need to widen the current 'name' paradigm to
>> include the notions of clan, kindreds, tribes, septs,
>> which prevailed before surnames came into being.
>> The current reporting tools, graphs, filters, reports, etc
>> in Gramps are directed mostly to 'surnames' (a mostly
>> Anglo-Norman construct) so this issue may be resolvable at
>> the reporting level rather than at the name level if there
>> is a good way to utilize the current "Name editor"
>> features to lay out the ontological scheme of an
>> individual . The reporting aspects would then only need to
>> reflect this available ontology. But I can't see a good
>> way. Maybe someone can.
>>
>> For example;
>> "Brían Óg mac Giolla Phádraig" (a patronym....sort
>> of)  was a 15th century ruler in what is now modern
>> Kilkenny.
>> "Dal Birn"  was his dynasty.
>> "Clann Conla" was his sept.
>> "Osraighe" was his tribe.
>>
>> Thousands of modern day people called FitzPatrick (a
>> surname) can trace back to him because it is well
>> documented that at a certain point the Dal Birn changed
>> over to using surnames and chose this Anglo-Norman one,
>> loosely translated from thiers; Fitzpatrick!
>>
>> From Gramps standpoint there may be an opportunity to
>> enable Gramps to record and report ancestry through these
>> tribes and septs when the individual surname ancestry
>> 'runs-out'. In reality switching over to surnames did not
>> signal a complete phase-out of the previous way they
>> thought of them selves and for centuries afterward you
>> would easily switch back and forth when someone asked your
>> name. It merely depended on who did the asking! So there
>> is a lot of available documentation too. This
>> name-plasticity and impenetrable nature of the the Irish
>> Language explain why Irish genealogy is such a hill to
>> climb. Many modern people of Irish descent, with the use
>> of emerging DNA tools, will soon be able to derive their
>> older sept and tribe affiliations as these become better
>> characterized. To be sure it will be very loose at the
>> begining and there may remain specific gaps in the
>> individual's ancestry but at least the paternal lines may
>> go back to these septs and tribes as haplotypes map closer
>> and closer to surname origins, septs and geographic
>> locations. We should think of this as the 21st century's
>> new wrinkle on genealogy.
>>
>> The question is how to accomplish this?
>>
>> In Ireland very many of these surname transitions were
>> well documented so we know (more or less) what modern
>> surname derived from what septs-tribe-patronym,
>> particularly if locations are known. This permits some
>> folks who can geo-locate their remote ancestor to 'bridge'
>> over that transition and continue backward some distance
>> in the tribal genealogies, which tend to be patronymic,
>> therefore with the addition of (arbitrarily chosen) 3
>> additional fields in the 'person' table of database and
>> configuring their alias in the name editor and a
>> user-supplied definition of the flow-down hierarchy
>> between them ("is a member of") Gramps might be uniquely
>> able to service this need by defining, for them at least,
>> the specific hierarchy. Whatever that definition was for
>> that individual would emerge in the reports, and could be
>> 'standardized' later merely by editing in the preferences
>> any new standardized labels.
>> "Preferences" Settings could provide the name aliases and
>> hierarchy for these three levels so in Ireland they might
>> be titled "sept-kindred-tribe", but in Montana for the
>> Plains Indians they could be called something else quite
>> different or in Russia or China something else. These
>> named patrilineal or even matrilineal lines would then
>> follow the individual backward through the named
>> ancestors. A problem might be seen where two utterly
>> distinct hierarchies mixed. A single set of settings may
>> not suffice for trans-tribal hierarchies. That would be
>> very rare.
>>
>> That leaves the reporting problem. The reports could read
>> the preference settings to provide options to display them
>> or report them or not. The 'familycolors' option would
>> merely select the type of descriptor to follow. Either
>> Tribe, Kindred, Primary Surname, Secondary Surname, etc.
>> In any case the data-links would be established in the
>> database for the named individual and would be available
>> for tool/report designers. There will emerge a wealth of
>> standalone genelogical tools which can consume a
>> standardized DNA data stream. I can think of a few
>> possible gramplets too.
>>
>> This ontological fabric may provide insights or even the
>> foundation for how to usefully encode, propagate and
>> report on genetic DNA results as well. Within broad
>> kindred groups patrilineal DNA haplotypes (the most
>> common) may allow for 'triangulation' and identification
>> of NPE (non parental events) which stump so many amateur
>> genealogists.
>>
>> The current wave of genetic genealogy increasingly popular
>> is quickly zeroing-in on the established population
>> dynamics and is revealing in medium to large scale what
>> families/tribes/septs/clans lived where while
>> simultaneouly revealing to an individual which of these
>> tribes his/her surname originated from. In my case this
>> has pointed me to two specific locations in Ireland, each
>> inhabited by a well known tribe carrying a single specific
>> haplotype which both they and I possess. These tribes
>> transitioned from patronyms to surnames between 11th-13th
>> centuries and have surnames which appear up and down my
>> own tree for centuries after. This has enabled me to spot
>> two (at least) potential NPE in my own tree. One through
>> Name change and another through adoption after second
>> marriage. Not everybody is comfortable identifying NPE in
>> their trees but Gramps users are pretty hardline
>> genealogists and can willingly consume the additional data
>> into their already well documented genealogies if they
>> want to. This will get easier to do as more people test,
>> so Gramps could provide a documentary platform for this.
>>
>> The third aspect of this would be the encoding in Gramps
>> of the genealogical data i.e. specific genetic mutational
>> signatures, STRs and SNPs. Each individual can/should
>> encode both the X and Y (if they have it) chromasome
>> markers. Done once for a tested user and the markers can
>> be automatically back-propagated with a click up the
>> paternal Y-tree or the maternal X-tree as 'presumed' data.
>>
>> For Y(male) data.
>> Gramps could display an additional tab in person view
>> displaying an array of tuples, locus name and value.
>> String and integer. A click on a + icon reveals a new
>> tuple. And so on until the user has created their STR
>> profile on the tab.
>> Due to the variable levels of STR testing available,
>> (anything from 12 to 500 sites on the chromosome)  Gramps
>> should adhere to the published Standard nomenclature for
>> these features and the user can select from a
>> pre-populated drop-down list for as many tested standard
>> loci that they know about from their test results. Each
>> Locus is described from the pull-down list as a
>> standardized text field (a label) and another Integer
>> field drop-down selects the value. Key-Value pairs.
>>
>> Example     Locus   Value
>>             DYS393  13
>>             DYS390  24
>>             DYS19   14
>>             etc
>>
>>
>> Similarly the terminal SNPs (if available from testing)
>> can be simply input in a user populated text field.
>> Private SNPs can be typed in to a series of additional
>> fields with a text field and pos. or neg. boolean assigned
>> to each private SNP. There would be many of these
>> initially and they might change over time as private SNPs
>> become 'officially recognized'. In this way negative
>> tested public SNPs can be encoded too. This last element
>> for the future tool designers.
>>
>> Example:    Public SNP              
>> Private SNP
>>             Type    
>> Value           Type    value
>>             R-P312  
>> +               abc     +
>>             R-L226  
>> +               def     +
>>             FGC5660 -
>>    Terminal R-M134  +
>>
>> X-testing (female MTDNA from both men and women) may be a
>> subset of these fields and could be on another tab in the
>> person view, or within the same tab but in a different
>> column. MtDNA Haplotypes vary so slowly over the centuries
>> that a single tuple would suffice.
>>
>> How would a user populate these two tabs.
>> (a) from tested certified lab-obtained values or (b) from
>> likely haplotypes obtained from surname/STR charting and
>> back-propagated SNP values from other testers.
>> Gramps users know a lot about their close and adjacent
>> family members situations so they can simply pull the
>> personal and modal haplotypes from the public databases of
>> their known tested relatives.
>> There would need to be a boolean for "tested", where
>> not-tested means 'presumed' because while it is likely
>> that your grandfather is the same as your father and
>> you....it is not certain, since randomly occuring
>> mutations and your grandma have a say in this matter.. But
>> all it takes is for your male-line cousin to test, and
>> verify to yours, and suddenly you are very confident that
>> the immediate male line is known!
>> This aspect may become very important in Tribal societies
>> with only recent surname adoption. Male lines verified in
>> this way can create kinship around shared ancestors.
>>
>> This is really just a little more genealogical metadata,
>> just like the citations, events, family definitions and
>> media we are already collecting. The next step, no more.
>>
>> The Gramps XML export could just encode these tuple lists.
>>
>> Exchanging your tree with someone not too (genetically)far
>> away from your close family would widen the triangulation
>> possibilities. 'Surname groups' can already consume
>> standardized pedigrees electronically and Gramps would be
>> able to interact with these communities of testers easier.
>> We might be able to crowd-source the haplotype of some
>> famous figures hundreds of years ago through this
>> triangulation.
>>
>>
>>
>> How could the data be stored?
>> I think lists of Name-Value tuples would be most efficient
>> for both STR data and SNP data.  Postgres supports this.
>> Other SQL based databases too, but not all. We do not need
>> to contain the raw data, the varient call files or the BAM
>> files. That is unnecessary for genealogy. The testing
>> organizations hold those.They are too big anyway.
>>
>> Anglo-Norman surnames are so 70's. ;-\
>>
>> I know that this is a volunteer project with restricted
>> bandwidth for the designers and coders, but this would be
>> a radical and unique genealogical feature.
>> Of course I may be dreaming and there is already something
>> available out there which I don't know about.
>>
>> I would welcome a discussion.
>
>
> On 11/23/2017 06:55 AM, Doug wrote:
>> Thanks for that clarification.
>>
>> Re your P.S.: I'd like to see your draft, as it may have
>> something in common with an old feature request for a
>> matrilineal filter.
>>
>> Happy Thanksgiving to you!
>>
>> Doug
>>
>>
>> On 22/11/17 19:19, brian fitzgerald wrote:
>>> I reported a bug 53133 and Paul Cully wgangled it. See
>>> below exerpt from the resolution.
>>> .
>>>
>>> .
>>> A NOTE has been added to this issue.
>>>
>>> ----------------------------------------------------------------------
>>>
>>>   (0053133) prculley (developer) - 21-Nov-2017 17:28
>>> Â  
>>> https://gramps-project.org/bugs/view.php?id=10279#c53133 
>>> ----------------------------------------------------------------------
>>>
>>> Thank you for the test tree.  It is quite a convoluted
>>> tree, and stressed out
>>> the deep relationship filter considerably more than the
>>> original author
>>> expected.
>>> I rewrote the filter to use a different algorithm and had
>>> it print out the
>>> longest 'path' it explored, turns out it was 675 persons
>>> long. The original
>>> algorithm broke at around 500 long.  A 'path' is from
>>> person to person through a
>>> family, looking for the connection between people.
>>>
>>> If you want to experiment with the fixed filter, let me
>>> know and I will tell you
>>> how to get and patch it in, tomorrow.  Otherwise you
>>> will have to wait for a
>>> newer version of Gramps.
>>>
>>> .
>>> .
>>>
>>> The posting to the forum was a bit more generic but
>>> related in the sense that I was not sure that the
>>> algorithm was trying to
>>> look both up toward a common ancestor and/also down
>>> toward a common descendant. Also there are two
>>> interesting edge cases in this algorithm
>>>
>>> 1) can the algorithm look for a pathway where there is
>>> both an up pivot individual (a common ancestor) and then
>>> back
>>> down the other leg of the  relationship to a second
>>> pivot individual and thence back up again. That is a lot
>>> of potential
>>> pathways to cover in a even a moderate size tree?
>>>
>>> 2) is a marriage without issue a valid relationship
>>> (think pivot point individual!) where the pathway can
>>> cross over? and if
>>> not can/should it be selectable as an option.
>>>
>>> In any case my Gramps was crashing at the first
>>> ultra-long pathway it found in my database(675 persons
>>> long) before it
>>> found some shorter more useful pathway so I never could
>>> test those edge cases.
>>>
>>> It also occured to me that I didn't know the name of an
>>> "inverse cousin", that is someone on
>>> the upside pathway to the common descendent. Idle
>>> curiosity really.
>>>
>>> I can wait for the next Gramps as Paul puts it. Should be
>>> long now until 5.0beta1, I hope.
>>>
>>>
>>> Thanks.
>>>
>>> Brian
>>>
>>>
>>> PS I have a draft posting for the forum which I am
>>> hesitating to send since it deals with
>>>
>>> Data and ontology structures which might require some
>>> more programmer/designer specific
>>> knowledge. It is a possible 'feature request' related to
>>> genealogical features found in tribes and
>>> kindreds. Probably an under-served user community.
>>>
>>> If I send it to you would you 'vet' it before I bother
>>> everyone. Else you can tear it apart
>>> in the Mantis Feature requests over the weekend. Happy
>>> Thanksgiving!
>>>
>>>
>>> Brian
>>>
>>>
>>> On Wed, Nov 22, 2017 at 11:41 AM, Doug
>>> <[hidden email]
>>> <mailto:[hidden email]>> wrote:
>>>
>>>     On 20/11/17 17:56, brian fitzgerald wrote:
>>>
>>>         I am not having any luck getting  
>>> a relationship
>>>         pathway displayed using rel_graph when the
>>>         relationship is not exclusively descendant
>>> on both
>>>         pathways. Better said, it seems to work OK
>>> when
>>>         there is a direct common ancestor (ie one
>>> common
>>>         pivot node person), but if one of the
>>> individuals
>>>         need to descend a couple of generations to
>>> arrive
>>>         where there is a common ancestor present
>>> (ie a
>>>         second pivot node in the pathway) (the
>>> pivot node
>>>         person is usually the result of  a
>>> marriage) it
>>>         fails.
>>>         The filters I am using  are the
>>> "pathway between
>>>         an individual and the result of a filter"
>>> and the
>>>         result of the filter is obtained by the
>>> simple
>>>         "descendant of an individual". I thought
>>> this was
>>>         going to achieve my goal but it crashed my
>>> Gramps. I
>>>         filed a bug report but I want to know if I
>>> am asking
>>>         Gramps to do something not designed in.
>>> Perhaps an
>>>         edge case which others don't need.
>>>
>>>         If someone has a  recipe using
>>> "filter math
>>>         magic" which they would share I would be
>>> grateful,
>>>         as might others.
>>>
>>>         Additionally, I suppose that two people
>>> connected by
>>>         a common descendant have a 'relationship'
>>> as well as
>>>         those who share an ancestor. I am not
>>> entirely sure
>>>         what you would call them, however. This
>>> aspect may
>>>         be useful moving forward as DNA admixtures
>>> need to
>>>         be tracked and triangulated, so Its not so
>>> much of
>>>         an edge case in reality.
>>>
>>>         Thanks
>>>          Brian
>>>
>>>     Can you give a couple of examples (anonymous) and say
>>>     exactly what means you're using to display a
>>> relationship?
>>>     Using Deep Connections I can't find any problem
>>> however
>>>     indirect the connection.
>>>
>>>     Doug
>>>
>>>
>>
>
>


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org
Reply | Threaded
Open this post in threaded view
|

Re: Showing relationship pathways between individuals

Nick Hall
On 29/11/17 13:40, Doug wrote:

> The third aspect of this would be the encoding in Gramps of the
> genealogical data i.e. specific genetic mutational signatures, STRs
> and SNPs. Each individual can/should encode both the X and Y (if they
> have it) chromasome markers. Done once for a tested user and the
> markers can be automatically back-propagated with a click up the
> paternal Y-tree or the maternal X-tree as 'presumed' data.
>
> For Y(male) data.
> Gramps could display an additional tab in person view displaying an
> array of tuples, locus name and value. String and integer. A click on
> a + icon reveals a new tuple. And so on until the user has created
> their STR profile on the tab.
> Due to the variable levels of STR testing available, (anything from 12
> to 500 sites on the chromosome)  Gramps should adhere to the
> published Standard nomenclature for these features and the user can
> select from a pre-populated drop-down list for as many tested standard
> loci that they know about from their test results. Each Locus is
> described from the pull-down list as a standardized text field (a
> label) and another Integer field drop-down selects the value.
> Key-Value pairs.
>
> Example     Locus   Value
>             DYS393  13
>             DYS390  24
>             DYS19   14
>             etc
>
>
> Similarly the terminal SNPs (if available from testing) can be simply
> input in a user populated text field. Private SNPs can be typed in to
> a series of additional fields with a text field and pos. or neg.
> boolean assigned to each private SNP. There would be many of these
> initially and they might change over time as private SNPs become
> 'officially recognized'. In this way negative tested public SNPs can
> be encoded too. This last element for the future tool designers.
>
> Example:    Public SNP              Private SNP
>             Type    Value          Type   
> value
>             R-P312  +             Â
> abc     +
>             R-L226  +             Â
> def     +
>             FGC5660 -
>    Terminal R-M134  +

Attributes can be used to store this information.

See:

8919: Add ability to record Genetic information eg: Haplogroup
https://gramps-project.org/bugs/view.php?id=8919


Nick.



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Gramps-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-users
https://gramps-project.org