Metadata should export with both sides of a relationship

[Faustyna]

If I have a Parents / Children relationship for Characters, and export it, we only get ‘Parents’ in that data. So all of the ‘Children’ characters have their parents listed, but none of those parents have any ‘Children’ listed. The column isn’t even there in the csv.

Even if I have a Romances/Romances bi-directional data point, it again defaults to just one side. If I export that data, it will show that John’s Romances include Jane - but Jane’s Romances will be blank.

Same thing if I try and tie together events and people. The People will have the Events in their column, but the Events will not mention the People.

I love using Aeon because it gives me the data from both sides and I don’t have to remember to type it in twice, but if it only shows up once in the export then I end up having to put it in again anyway.

I asked about something similar almost two years ago in regards to Scrivener. Only now I realize it doesn’t work for bi-directional relationships either if I’m just looking a the csv data.

Please allow Aeon to export with both sides of a relationship being accounted for.

[Amontillado]

The information you want is there, if you read between the lines, and this might be fixable. I can give you an Apple Numbers (Mac spreadsheet) solution which I believe could be applied in an Excel pivot table.

I opened an Aeon CSV export in Numbers and turned on Categories (like an Excel pivot table, except live-updates). I categorized by value on the Parent column.

The rows starting with inverted triangles are parents. The intervening rows list the children of that pair.

Is that what you need? For an illustration of how data is agnostic to marriage vows, I added a character named Bubbasue Linton, the unfortunate issue of a rave attended by Heathcliff Linton and Mrs Earnshaw. No disrespect intended, it’s just an example. It is my loss I’ve never read Wuthering Heights.

The “blank” category shows characters who do not list parents.

Apologies for an Apple-only example. It’s what I use.

If Excel won’t do the job, Easy Data Transform is a data manipulation tool that I get enormous value from. It will do pivot and un-pivot tables, along with many other operations.

For brevity, I hid most of the columns in the Aeon output. Every column is still there in the Numbers categorization table.

[Faustyna]

Thank you! I’m on Windows but will look into alternate programs like that. I didn’t know they existed

[Amontillado]

Easy Data Transform (EDT) is magic stuff. I have no affiliation with the developer (but wish I did).

Here’s another spreadsheet I created from Aeon, this time an Excel spreadsheet without using any Apple specific software:

Here is the actual EDT “program” used to walk the Aeon export into that spreadsheet. It’s just a few little functions, all of which are standard EDT operations, linked together graphically without any coding:

The pink box at the head of the chain means “read a file called wuthering.csv,” which is the untouched output from Aeon.

That file flows into a filter, which I set to keep rows with type Character and also had something in the second Parent column. The first Parent column in the output was Aeon’s parent event link. There is also a Parent relationship type. That’s the one I wanted to use.

But there was a problem. Isabella was the daughter of “Mr Linton,Mrs Linton.” Edgar’s parents were listed in reverse order, “Mrs Linton,Mr Linton.” We need the parents to be listed the same.

Split Col split the Parent column on commas, producing as many columns as necessary to accomodate as many parents as any character had. In this Victorian setting, two columns were all that needed to be created, Parent 1 and Parent 2.

I had Split Col sort the component parents alphabetically. That let Concat Cols reassemble the parents consistently.

Remove Cols tidied up. It nuked every column except the Label column, which contains the children’s names, and the Parent 1+Parent 2 column created by Concat Cols.

Sort re-ordered the rows by the concatenated parents. Rename Cols gave the columns better names, and the last green bubble wrote the table out as an Excel file.

EDT is a great companion for any workflow involving data tables, JSON files, spreadsheet, CSV/TSV files, XML, fixed record length files, and probably a few formats I’m forgetting about.

[Faustyna]

Wow, thank you! Looks like EDT is a paid program (though it has a free trial), so I will attempt to make this work in Excel or LibreOffice first.

I wonder if I could do more with this too, like put the children in order based on ‘Start-Date’ and therefore have them listed from oldest to youngest. It could revolutionize how I deal with all of this info, and finally let me have a database/wiki that is easy to look up things in, without me having to triple-check if I put in the right info in 4 different places…

[Amontillado]

Absolutely. What you sort on is totally up to you.

If there is something needed that EDT doesn’t support out of the box, you can add Javascript steps. The coding is very simple, usually an equation or two and you don’t need more than a brief intro to Javascript.

I had the high honor of bringing to light some economic hijinks in a backwater government office. EDT was a siege weapon in that deal.

Aeon Timeline was also part of that. I used it for tracking correspondence, who I had mailed/emailed/spoken with and the return contacts. It also helped because keeping track of how stories changed over time was part of the dustup.

Correspondence and notes went in Devonthink. Events in Aeon had links to details and logs in Devonthink. Strong tools, flying in close formation. Government uses “big data” - and so do I, at least in my own small scale way.

[Faustyna]

Oh that’s good to know! Thank you. I like how EDT works and looks, so I may end up buying it. I’ll look into javascript too.

Right now I’m still figuring it all out. I wonder if EDT could do something like - if {Label} is in Column “Parents” for a certain Row, add {Row Label} into column “Children” for {Label}.

Or more complicated: If Label (say Elizabeth) is in column ‘Relevant’ (for, say, an event like Pierre and Elizabeth marry) add new column ‘Important Dates’ to ‘Elizabeth’ with “Pierre and Elizabeth marry: Start Date of Pierre and Elizabeth marry.”

I assume that might be possible with javascript

[Faustyna]

Still haven’t managed to make EDT work the way I want it to to do this. Or excel, or libreoffice calc. Also trying to come up with maybe a python script that can fill in the blanks.

It would be a lot easier if Aeon did export both sides

[glimfeather]

@Amontillado

Easy Data Transform (EDT) is magic stuff.

My Aeon usage does not need such a tool but my Apple Health data does. Each month I export that data and at the start of this month (May 2024) the unzipped XML file was 225Mb. Looking for easy way(s) to convert that to something more useful for analysis (such as CSV for import into the stand-alone SQLite3 executable on macOS or maybe pandas.)

[Peter_T]

If you export the aeon file to Obsidian, you can have the backlinks displayed.

[Amontillado]

That’s the thing about tools like EDT. It’s an intermediate tool between data and your end use and it’s very capable. I’ve sliced through data sets as large as 600,000 lines of fixed record length data, 4000+ characters per line.

Imagine you want to analyze books for a government office. They have to cough up their records but they don’t need to make it easy. If you can get a CSV dump from their accounting system, you can browse and convert to other formats.

For instance, accounting records, perhaps filtered by some criteria, to Aeon Timeline. You can do more than just follow the money, you can seize it by the lapels and shake the truth out of it.

[glimfeather]

@Amontillado

I’ve sliced through data sets as large as 600,000 lines of fixed record length data, 4000+ characters per line.

My current export.xml from Apple Health has 1.2million lines of variable length some of which are discreet data points and others are correlations (Apple’s term for related entries) such as Systopic and Diastolic blood pressure readings. At some point Apple changed the way the data was exported with discrete Systolic and Diastolic reading and then later “correlated”.

Of course this need is not directly related to Aeon but your investigative accounting suggests that putting Apple Health data into a Timeline might reveal correlations between other data items.

[Faustyna]

I still have a few days of the trial, so I’ll keep at it. I suppose I just wish it were easier, since I plan to do a lot of worldbuilding in Aeon and then export it to Obsidian. But I also plan to keep updating the information and changing it as things happen- meaning I would have to sift through the data and rearrange things every time I changed something. Making it a hassle if it’s not fairly quick.

I really do hope Aeon eventually syncs with YAML/Obsidian. The ‘quick workaround’ I was trying to make turned into a monster of a project… ironically by the time I finish it, watch the sync be supported