Anglo saxon kinship

[stschiff]

this is a branch that I recreated after prematurely having merged #136. I will likely merge #104 soon, but more work is needed on this PR to get it done.

[stschiff]

This is on my desk to review for now.

[stschiff]

OK, so I've now gone through Table S1 of Gretzinger et al. 2022 and added all relationships except for Buckland (BUK). @93Boy, please go through BUK and add the relationships, you can take a look at the other sites how I have encoded things. here are a few best-practices that I've found:

1. I go through the Table S1 one by one and add the respective relationships to the janno file, in the order given in the table, not in the Janno
2. Joscha has entered reciprocal relationships for all pairs, so you won't miss anything if you strictly go individual by individual, no need to jump and add the reciprocal one immediately, but you'll get to it when going through the table.
3. Joscha has often written "second to third". In these cases I have opted to use "second" as Relation_Degree and then added to Relation_Note something like "second_to_third".
4. Remember that all Relation_* fields are list columns, and the order matters. For example, if you have a note on the second relationship, but no note on the first, you'll need to write something like _;note_for_second in Relation_Note.

Let me know if you run into problems. Otherwise please finalise the BUK relationships and mark this PR as "Ready for review" when you're done. Thanks!

[93Boy]

@stschiff I have added the BUK relationships but when I try to rectify the file afterward, trident throws this error Can't read sample in ./2022_Gretzinger_AngloSaxons.janno in line 48: parse error (Failed reading: conversion error: expected Double, got "The Bishops Seat" (Failed reading: takeWhile1)) These fields do not relevant to relationship data.

[stschiff]

I'll wait for your green light to review this again, @93Boy. Or just click "Ready for review", please

[93Boy]

I have gone through multiple consistency issues in this file and I have noticed the file structure is corrupted and data fields have been severely mixed up. I tried to retrieve the data from the master branch but seems like the master branch also has the same mixed-up file structure. I appreciate you having a look and giving a guideline. Another technical issue is you have categorized several relationships as a third degree but trident support only identical, first, second, thirdToFifth, sixthToTenth, other relationship types. But this issue can be fixed within a few minutes. My major concern is about the corrupted file structure

[stschiff]

OK I will take a look whether I can see what's going on. Regarding the third-degree relationships: We discussed this in the meeting: You should then indeed set Relation_Degree "thirdToFifth", but you can use Relation_Note to then say "third degree"

[stschiff]

OK, @93Boy, I have now fixed the structural issue. When you run trident validate on the package, you will see

[Error]   Can't read sample in 2022_Gretzinger_AngloSaxons/2022_Gretzinger_AngloSaxons.janno in line 21: The Relation_* columns are not consistent
[Error]   Can't read sample in 2022_Gretzinger_AngloSaxons/2022_Gretzinger_AngloSaxons.janno in line 89: The Relation_* columns are not consistent
[Error]   Can't read sample in 2022_Gretzinger_AngloSaxons/2022_Gretzinger_AngloSaxons.janno in line 90: The Relation_* columns are not consistent
[Error]   Can't read sample in 2022_Gretzinger_AngloSaxons/2022_Gretzinger_AngloSaxons.janno in line 93: parse error (Failed reading: conversion error: Relation degree "third" not in [identical, first, second, thirdToFifth, sixthToTenth, other])
[Error]   Can't read sample in 2022_Gretzinger_AngloSaxons/2022_Gretzinger_AngloSaxons.janno in line 94: parse error (Failed reading: conversion error: Relation degree "third" not in [identical, first, second, thirdToFifth, sixthToTenth, other])

These issues are caused by inconsistencies between the Relationship-Columns, where you have different number of list-items in the different columns. Please take a look and fix.

[stschiff]

Reminder, that this is on your desk, @93Boy, unless I'm told otherwise. Thanks.

[stschiff]

OK, so this looks like it is ready. I can't approve my own PR, @AyGhal and @nevrome, so feel free to approve and ultimately merge this. Thank you @93Boy

[nevrome]

Cool that this is ready now! Very impressive!

I only see some details:

• The package version now jumps from v1.0.0 to v1.2.0. I think v1.1.0 would be a bit more consistent.
• The changelog message "Relationship_update" is pretty short and technical for such an amazing piece of work.
• This package is actually broken, there are a handful of these warnings

[Warning] [2024-03-08 08:31:05] Issue in data/2022_Gretzinger_AngloSaxons/2022_Gretzinger_AngloSaxons.janno in line 23 (Poseidon_ID: HID001.A): Date_Type is "C14", but either Date_C14_Uncal_BP or Date_C14_Uncal_BP_Err are empty

This PR may be a good place to quickly fix this. In this case the version could be incremented to v1.1.1, I think.

[stschiff]

What the... I actually ran trident validate. How could I overlook this. OK, I'll take care of this.

[stschiff]

OK, I've fixed all the points raised by @nevrome except for the broken C14 lines which I'm gonna fix now as well.

[stschiff]

OK, done. I've fixed all the errors. I will add one issue, as the C14-data here can't be quite correct for one line. But the validation should now pass.

[nevrome]

Looks good to me now. Can be merged 👍