Common features within collection of tiddlers

[Bill Bell]

Can I identify a collection of Tiddlers within a wiki and have TiddlyWiki tell me which features or characteristics are shared by those Tiddlers?

[Mark S.]

What kind of features are you thinking of?

[Charlie Veniot]

Although I say ABSOLUTELY (!!!), what I have in mind involves a horrid mess of filters in a process that needs to know each and every feature/characteristic to consider.

It ain't pretty if features/characteristics are few and simple, and it is downright ugly otherwise.

[Bill Bell]

Suppose that the Tiddlers represent individual people, and that within each of these Tiddlers there are facts about those people that they have in common. The facts could be represented in any way available in TiddlyWiki. For instance, a fact might be a person's occupation or birth year or favourite author.

[Charlie Veniot]

Say you are interested in the collection of people who are librarians, and you want all of these librarians grouped by those features/characteristics.

Do you want the results to look like:

• 10 librarians were born in 1965, and favourite author is Margaret Atwood
• 2 librarians were born in 1970, and favourite author is J.R. Tolkien
• 5 librarians were born in 1971, and favourite author is Margaret Atwood
• 7 librarians were born in 1971, and favourite author is J.R. Tolkien
• 3 librarians were born in 1982, and favourite author is Agatha Christie
• ....

If so, that can certainly be done.

[Bill Bell]

Thank you, Charlie. But, alas, no.

I would like to select, say, m people from a TiddlyWiki (somehow) then have the TiddlyWiki tell me what common features these people have. 

[Mark S.]

Tiddlers can have fields. So you can assign fields for birth year, favourite author, occupation, etc. As in all databases, it's important that the input information be consistent. So you wouldn't want a person's occupation to be input variously as Doctor, Physician, Radiologist. Or a favorite author to sometimes be Tolstoy and others as "tolstoy". Or "Toystory". You can make a template to help with tiddler creation,  so that when you match by favorite authors, it will always be by "Tolstoy" .

[Charlie Veniot]

Can still be done.  A bit complicated to build that kind of filtering smarts, but I think that would be a really fun project to work on for somebody who has the time.

If the result is a simple count of tiddlers for each individual tag, and a simple count of tiddlers for each possible value in a field (that done for each field), that would be easier than getting counts for each possible combination of tags + each possible combination of field values.

And the process does not need to be told what tags exist, what fields exist, nor what field values exist.  The process can figure each of those out automagically, with maybe just a hard-coded "exclude" list if there are certain kinds of tags, and/or fields, and/or field values that should be excluded from the process.

[Bill Bell]

Thank you both. What I conclude from this is that I need to learn (quite a bit) more about TiddlyWiki — and then all of this will become clear to me.

[Charlie Veniot]

Although I may be a cheap date ...

On a whim and for the $hit$ and the giggle$: 

I am currently looking for employment (my online curriculum vitae à la TiddlyWiki).  Little and fun side-gigs, enough to keep me afloat, are certainly welcome.  Contact me privately via my contact form to discuss.

So if of any value to the community and with adequate funding (that would be the not-so-cheap-part o' me), I can build an open-source something generic enough that can handle this kind of purpose anywhere for anybody.

BTW: I'm kind of big on the idea of "release early and release often"  (i.e. proof of regular progress) and the ability for any party to pull the plug at any time for whatever reason, for example: change of scope/requirements, progress too slow, solution too complicated, complete and utter failure, too many hugs from me ...  Some sort of milestone basis.

[Mat]

bell...@gmail.com wrote:

 

I would like to select, say, m people from a TiddlyWiki (somehow) then have the TiddlyWiki tell me what common features these people have. 

Do you mean without specifying what exact "common features" you are looking for? What defines a "feature"? If you stick to well defined "features" that you can actually refer to, then sure, it can be done. If it is just any unspecified feature, the question is probably not meaningful.

<:-)

[Charlie Veniot]

It really should be no problem, albeit a bit or work, to create a generic process that can report: "the list of commonalities in a specified set of tiddlers."

The process wouldn't need to know anything at all about what tags or what fields to consider.  It could simply consider all tags and all fields and all field values.

A bonus would be to allow some sort of inclusion or exclusion list: to say "just consider these features", or to say "all features except these."  And/or a way to hide any reported item à la wysisyg as one is looking at the results of the query.  So click on any little associated reveal widget "not interested in this result" button.

That would be pretty cool.

Aside old guy reminiscing

Many moons ago in my CS "Discrete Structures" class, the prof gave us the bonus assignment "write a proof that the following is impossible to do or, alternatively, write the program to prove it can be done:  is it possible for a program, with no input, to contain itself in the program and output itself upon execution."  I thought: "hey, that's what a virus does."

One guy in the class wrote the proof that it could not be done.  He did a fantastic job.

I was the lone goof who wrote the program during an all-nighter in three large lines of GW-Basic code.

We both got equal bonus marks.  I probably should have learned something from that ...

[Charlie Veniot]

G'day,

I felt like playing a little bit, and quickly put together the following very basic tiddler (drag to any TiddlyWiki and import it to try it out): Tell Me About My Tiddlers.

The tiddler has two "report" sections:

• list of all tags and count of tiddlers that have each tag
• list of all fields and count of tiddlers that have each field
  • AND, list of distinct values per field along and count of tiddlers that have each value

So just a little show'n tell about the potential to put together a "Common features" query/report.

On Sunday, January 31, 2021 at 4:19:02 PM UTC-4 bell...@gmail.com wrote: