Bring back the author nationalities!

See http://www.librarything.com/topic/144601

There used to be a list of authors by nationality on http://www.librarything.com/profile/lorax/stats/nationality which currently just shows a pie chart of countries representing more than some fraction of your library. So while all the data is there in CK, there's no way to see "Which countries do I have authors from" or "Which of my authors are from New Zealand".

This RSI is to bring that list back; we've been clamoring for it for years on that thread. The pony version would be to get a map. The same Google API they're using can do it (or it could, a couple years ago when I last used it. I haven't looked at it for a while.)

+1

Yes, please bring back the list.

Bump.

If one of my RSIs makes the Great List, this would be my choice.

Bump.

I never got official notification to shut up about this for a year (i.e. the "Red" status), so I'm bumping it.

Bump.

There are a bunch of us who would like to see this resumed someway.

Here here!

There there!

Yes please! >1 lorax: everyone should know how many New Zealand authors they've read. ;-)

Bump.

Just watched TED talk about woman that read a book from every country recognized by the UN. Would like to do that. Need to know which country our books/authors come from. Tried settings and "where books are from" only yielded the publisher--not even a city on that option.

You can get author nationality on the author page CK, but there's no way to even get that in your catalog (which I would accept as an alternative).

I do it myself with tags, since there's nothing else.

>10 Lit-with-Wit: Yes, it was a good talk. An excellent way to experience other cultures and perspectives without ever having to leave your living room. Some reader sites and blogs encourage more diversity in our reading, would be helpful of LT to help with this.

>12 .Monkey.: so does it work for you? Any suggestions in formatting the tags?

I mean it's not the same as having it properly in the details, but at least it puts the info at your fingertips. You can search for it, you can click it and see the books tagged with it, it's useful. I simply do a "nationality: DEMONYM" (e.g. nationality: Russian) setup, or well, I did before I switched to using a symbol instead of the word, but same difference. If you prefer you could just do the country name, rather than demonym, but I also put the country setting in tags so I'd rather the two don't mix.

>14 .Monkey.: thanks, might try it, I know I don't read widely, that's why I'd like the nationality list back, and/or an improved chart.

Yeah, if you group them with some sort of header (using "nationality" or whathaveyou and/or putting some kind of symbol in front) then they appear all together in the tag list, so it's easy to skim over it and see what the numbers are. :)

Bump.

Bump.

Please!

Where where!

(In reference to >7 krazy4katz:)

Awww…. :-)

Bump.

Bump.

Ooh yes please please bring this back. I'd gladly do without being told that my books would make a stack as high as the tower of Babylon and being reminded of the names of characters in my books and of groups that a programmer thinks I might want to join if I could have this feature instead. (In fact I'd very gladly do without those things regardless, but my point stands.)

BUmp.

Bump.

Bump.

Bump.

Bump.

Bump. Not gonna let this go.

Bumping

Time for another bump, I think. So tired of seeing "US, UK, Canada, Australia, Other" with no way short of clicking through to every single author in my catalog of finding out what that "Other" represents.

Bump. How, exactly, is this useful or interesting? All the data is there. Why can't we see it, either here or by having "Author nationality" as an available field in our catalogs?

And here, on the other hand, is what we COULD easily have:



Generating this was a tedious and manual process because I had to individually visit 1000+ author pages to grab the nationalities - something that would be a trivial join for LT, that they're already doing anyway to make the stupid useless pie chart. Once I had the nationalities, it was the work of a moment to do the aggregation and use Google's API to make the chart (plus another minute to decide that my author list was heavily enough dominated by US + UK that using a log scale for the colors looked better.)

>35 lorax: Now, that's assiduity. Fair play. How long did it take you to create this?

I don't think the pie chart is useless. I just don't think it's enough.

I should have posted this in a month's time, shouldn't I, to give the thread another bump then.

I don't remember if this was discussed before, but how is this supposed to work for authors with multiple or ambiguous or contested etc. nationalities?

...or if the nation of the author's birth no longer exists.

>37 LolaWalser: The system went by the last entry in the list under "Nationality"-- I think the presumption being that ideally you'd list them in chronological order.

>38 jjwilson61: There still remains a CK field on each author page called "Country (for map)" that exists for this express purpose. (So you could enter, say, "Italy" for Ovid.)

>39 Stevil2001:

The system went by the last entry in the list under "Nationality"-- I think the presumption being that ideally you'd list them in chronological order.

That makes no sense whatsoever to me. Lots of people can lay claim to multiple nationalities at the same time.

Not to mention that "last entered on LT" doesn't have to refer to some biographical sequence of events at all and can be purely accidental.

(So you could enter, say, "Italy" for Ovid.)

I have to say, I hope no one would do any such thing. Italy came into existence in 19th century, Ovid was a citizen of the Roman Empire. Anachronisms of this kind just feed chauvinistic megalomania.

Dammit, the "ancient authors" is going to scupper this again, isn't it, like it did last time (when there were dozens and dozens of posts arguing about it, which is why the "country for the map" field got created)? Despite there being a "Country for the map" field, separate from Nationality, that could be blanked out in those cases, and despite many people having libraries where those would only affect a handful of authors?

Damn, damn, *damn*.

Look, if the map truly is a non-starter, can we at least get the list? It's just exposing data that's already there. And if I can download the list, I can make my own damn map.

(Anybody can! It's a chart option in Google Sheets! No coding required!)

Well, for my part I'm interested in the reasoning behind whatever gets implemented, not in "scuppering" anything.

If this "country for the map" thing refers to currently existing countries and not to various historical countries pertinent to the author in question, then obviously no, it will not do for many authors. It does not do to claim Ovid/Ovid's nationality as Italian or St. Jerome as a Croat, no.

But that is still only one sort of situation. And easily remedied by adding countries which no longer exist. Or marking CLEARLY that the connection is merely geographical and not cultural.

But, again, what with those with multiple nationalities? What is even meant here by "nationality" and how is this identification different from ethnicity regarding this LT stat?

Take British authors for instance--do we want them all together under that rubric, or is the point to differentiate between English, Irish etc.?

Of course, there is always "Other/None of the above". :)

Guaranteed to be MY favourite category...

>38 jjwilson61: That's why the "Country for the Map" field was created

LolaWalser (#44):

I encourage you, then, to peruse

https://www.librarything.com/topic/144601 and its predecessor https://www.librarything.com/topic/104017 (the screenshot at the top of that one shows what the graph used to be, which is far better than it is now, including lists of authors from each nationality.) The distinction between "nationality" (potentially multi-valued, cultural, and not necessarily corresponding to a currently recognized sovereign state) and "country for the map" (single-valued and geographical) was made sufficiently clear there that I unfairly assumed anyone in this thread was up on it and continuing an old argument rather than asking about the rationale for the current setup.

bluepiano (#36):

It took a long, long time to actually get the list of author nationalities - I did it a bit at a time over the course of a few weeks. Once I had those in a spreadsheet, it was a quick cut-and-paste into the example GeoChart API that Google has as an example to turn it into a map, or to do the same with the built-in map chart that Google Sheets offers. Plus another minute to convert the counts to a log scale for more nuance in the shading. Definitely under five minutes from spreadsheet to map, though! Really, if they would just give us the data in an exportable form, I'd be happy.

>47 lorax:

Thanks for the links. I take it the presence of "Rome" and/or "Roman" in that pie chart indicates that historical countries pertinent to specific authors were to be included, so that's good. Included, that is, in some list or graph or chart but not a geographical map--which makes sense to me but perhaps for wrong reasons (I imagine a geographical map rendered useless by a zillion superimposing, crossing outlines or some such...)

Other than that--and now I'm just saying generally, lorax, not addressing you personally--having skimmed the thread top to bottom, I see nothing but problems--not for ALL authors, to be sure, maybe not even a large majority, but a sufficient number that I'm not sure I'd find this feature anything other than infuriating.

One size does not fit all. There is no category "nationality" that can be pinned down with a singular descriptor for ALL LT authors. As long as the data must be represented by singular choices, it will be wrong for masses of authors.

But in addition, and this may be the worst aspect of it, it will be wrong internally, because even the single choices made here do not mean the same thing concerning all the authors. Sometimes they refer to citizenship, sometimes to ethnicity, or even just language.

It makes me wonder again just what is the point of it. I genuinely don't get it. What is the question this summing up, whether in a list or chart or map, is supposed to answer?

"What passport did this person have?"

"Where is this person from?"

"What is this person's parental/maternal/paternal language?"

"What is this person's ethnicity?"

There do exist people for whom ALL the answers here are identical and unambiguous, of course (at least in some specified period), but, in my opinion at least, there is no point whatsoever in combining such cases with those for whom the answers are not identical and unambiguous.

OK, I guess I worked down to the rock bottom of the issue for me--I see the point of having a chart/map/whatever for the incontestable, unambiguous ones, for those that can go as "like with like", for those for whom nationality means the same thing. But I'd leave everyone else out.

>49 LolaWalser: I don't get what's infuriating about a feature you don't want or need, but others do. It's not as if it will open itself and harass you with a map. You'll simply ignore it as not useful or informative.

Personally, I'd love too see the geographical spread of my books. It seems like a minor project to make it work, but then again I'm not a programer

>50 SandraArdnas:

First, you are quite right that I can and would ignore it, that's not a problem at all. But that doesn't mean I can't have a conversation about what the feature is supposed to do and whether it answers that purpose.

Why would it/could it be infuriating--because bad data is infuriating in itself; because errors ought not to be reinforced and propagated; because nationality and ethnicity are often as volatile a topic as they are ambiguous. In short, for quite a number of serious reasons.

Personally, I'd love too see the geographical spread of my books.

OK, but, if you don't mind explaining, what does "geographical spread" mean here? Is it one thing only for all your books, comparable within that set?

What would it mean, in your view, for authors like Nabokov or Shaw or Henry James and similar?

LolaWalser:

You do realize, of course, that all the data, ambiguous and confusing and controversial, is already there. If anything, making it more prominent might encourage people to correct problems - to enter additional nationalities where they're absent, to correct those that are blatantly wrong, and so forth. Or are you suggesting that increased visibility of the data would encourage people to thoughtlessly enter inaccurate data - "Italy" for all ancient Roman authors in "Country for the map", for instance? That's possible, I suppose, but I'm willing to take the chance.

>51 LolaWalser: It would mean whatever people put in the Country (for map) field. This approach divides the issues of implementing a desired feature and debates about particular individual disputable cases ;) The latter are not really tied to the feature itself. People can discuss and decide on individual cases now, as well as after the implementation.

I have no idea why Nabokov, Shaw or Henry James would be particularly disputable. I consider both Nabokov and James American authors, but if the consensus were to be different, I'd be outvoted on that particular case and they would be charted in line with the majority opinion. Kafka I find more puzzling where to chart, but I'd like to repeat, if the feature requires one particular geographical location, than we eventually decide on one and put in the country-map field. The nationality field itself still allows multiple data rows when needed, so that data will not go anywhere whether we are talking about dual citizenship or long gone empires.

>52 lorax:

True, various kinds of errors are already there, but if I understand correctly, the data is in individual CKs, not aggregated, and the CKs can be enriched, whereas expressing "nationality" in a graph/chart/map is of necessity (?) reductive to one choice. It's not necessarily the same kind of ambiguity or level of error, is it?

For example, if the CK for Nabokov says "Russian, French, American, Swiss" (haven't checked, just as an example), that's ambiguous all right, but it needn't be wrong. For him, all those descriptors are/may be valid in some sense.

But to reduce that complexity to just one of those creates an error that can't be corrected by replacing it with another of those terms.

And again, just saying here, I've no intention to take away or ruin anything for anyone. If the heterogeneous nature of this data doesn't upset people, then it doesn't--it's not a hill I'll choose to die on any time soon! ;)

>53 SandraArdnas:

I consider both Nabokov and James American authors, but if the consensus were to be different, I'd be outvoted on that particular case and they would be charted in line with the majority opinion.

Personally, I don't accept that a "majority opinion" is necessarily worth bollocks in discerning matters such as these. But I don't disagree that it would work as a blunt instrument to push through a singular choice necessary for a one-on-one mapping. I still don't see what the value of the latter would be, what information it's meant to communicate... but I guess there's no point in repeating my questions.

However, the idea that it would be acceptable to think of Nabokov as American tout court made me wonder about something else--what about a chart/map/etc. based on the original language of the individual titles? Wouldn't that create a "geographical spread" more readily relevant to one's books than the authors' "nationality"? Language is less ambiguous than "nationality".

I'm not thinking either/or here, suggesting a replacement of one data set by another. If there's one chart why not twelve. I'm just still trying to understand, as above, what question this map is supposed to answer, and whether, for some people/situations at least, original language doesn't answer it better than "nationality".

Language and cultural affiliation is what makes me consider Nabokov an American author. As for mapping according to language, that info is already available under stats/memes, both according to the language of the edition and original language. It is also not very informative geographically. Half the world publishes in English, and the remaining half in 5-6 other languages.

However, the idea that it would be acceptable to think of Nabokov as American tout court made me wonder about something else--what about a chart/map/etc. based on the original language of the individual titles? Wouldn't that create a "geographical spread" more readily relevant to one's books than the authors' "nationality"?

That would work pretty well for European authors, I think, but you'll get most African authors writing in either English or French, have most of Latin America lumped together as writing in Spanish, and have lots of authors choosing to write in English to get the larger reader base even if it isn't their preferred language.

>53 SandraArdnas: I consider both Nabokov and James American authors, but if the consensus were to be different, I'd be outvoted on that particular case and they would be charted in line with the majority opinion.

I really doubt that you'd get Tim to add a voting feature here, so what you'd get wouldn't be majority opinion but the opinion of the last person to care enough to set it.

Language and cultural affiliation is what makes me consider Nabokov an American author.

Funny, language and cultural affiliation make me NOT think of Nabokov as an American author. I'd think his Russian work is at least as important as the English-language output, and as for affiliating to "culture", he ended up in the States to escape the war and skedaddled back to Europe as soon as he had the money to do so.

It is also not very informative geographically.

True. However, neither are all authors' nationalities--especially when reduced to a single choice.

Half the world publishes in English, and the remaining half in 5-6 other languages.

And yet for some reason even the short list of languages on the drop-down menu in our catalogues here counts more than double that.

I guess we're done here.

>57 lorax:

Yep, agreed, a map of original language would not work. That's why I said chart--like the pie chart in the other thread... if the visual was the important feature (vs. a dull list or graph).

>55 LolaWalser:, >56 SandraArdnas:, >57 lorax: Then you get situations where Latin was the original language - which covered a large geographic area and a massive timespan. Do you break that into Classical Latin, Late Imperial Latin, Medieval Latin and so on?

>60 Maddz:

I don't think lorax will thank me if the discussion diverts to original language. ;) I only asked that trying to understand what question this feature is expected to answer. And as I mentioned above, I did not envision a map for original language, I specified "chart".

>58 jjwilson61: Yes, I didn't think we would really vote, but rather discuss the most disputable cases, whereby those in minority would accept the predominant opinion and not wage edit wars

>59 LolaWalser: I'm then ignorant of a good chunk of his work. I'm aware how he ended up in the US, but in my mind his major works are in English and as an American citizen. But either way, this is something decided on a case by case basis. If you need to pick one primary affiliation, you pick one that makes the most sense and the more detailed and thus more correct data remains where such ambiguity is possible, which in this case is the nationality field.

As for language map, I meant to point out that a select few languages cover the majority of the globe, not that there's a scarcity of languages as such.

I'm confused now -- what are we arguing about?

Is it the "Country (for map)" field? Is it the attempts at visualising author nationalities? Is it the particular visualisation of a world map that is the issue?

What, specifically, is the problem?

>62 SandraArdnas:

As for language map, I meant to point out that a select few languages cover the majority of the globe, not that there's a scarcity of languages as such.

Er, what? You said that "Half the world publishes in English, and the remaining half in 5-6 other languages.", which patently isn't true. The world publishes in hundreds of languages; how much of that is reflected in anyone's personal catalogue is another matter.

And for the third time, I WASN'T talking about getting the original language MAP, but a (PIE) CHART. See the second link in >47 lorax: for an example.

Geographically for heaven sake. Look at the globe and the spread of languages geographically. The RSI is about a map

Okay, can we please get this back to the original RSI? If someone wants to make another RSI for a pie chart for original language, I personally wouldn't find that nearly as interesting (for the reasons I outlined in #57: I think it's interesting to know if someone is from Nigeria or New Zealand or Barbados and writing in English rather than having them all just collapsed into "English"), but wouldn't be opposed.

I want the data that is currently in CK in "Country for the map" or "Nationality" to be exposed in a more useful and accessible manner than the current pie chart, which shows only the top few, lumps all the rest into "Other", and has no listing of names corresponding to the different countries. This could be:

* A restoration of the original pie chart which included a listing of which authors were in which category, including an expansion of those combined into "Other".

* A map, preferably with the list of which authors were associated with each country, a mockup of which is posted in #35.

* Just making the data available in tabular form via a meme or an exportable catalog column, so that those of us who want can create our own visualization.

>65 SandraArdnas:

Yes. I KNOW. I asked the bloomin' question (in this bloomin' post >55 LolaWalser:, see for bloomin' context) about bloomin' original language trying to get a hang on what question you bloomin' think this feature answers. Because if you think the answer's "nationality", it bloomin' ain't. The data is ratty and the fucking resulting MAP a hundred times worse.

Now maybe we can break off this "dialogue" for good, because not only you don't seem to be reading what I post (not reading for comprehension anyway), but refuse to acknowledge even what YOU posted.

>66 lorax:

What's your solution for representing authors with multiple nationalities, or ambiguous/contested nationalities, and those who were nationals of countries that no longer exist?

I think the first step to introducing some clarity in this utter mess of a feature is to unyoke it from the idea of "nationality".

How about creating a new category in the CK specifically as a source for the map? A drawer with a label that expresses something along the lines of "the country with which this author is geographically/historically/most readily/likely/often/whatever associated with".

>68 LolaWalser:
"How about creating a new category in the CK specifically as a source for the map"

CK has is a "Nationality" field, where you can put any former country-like entity, any combination of ethnicities and state affiliation that you please. There is also a "Country (for map)" field. If there is going to be a map based off CK nationality/geography data, then that is the field that's going to be used.

I know you know this. I don't know why you ask a question you already know the answer to. If this were not LT, and if you hadn't been a long-standing member, I'd mentally categorise you as a troll, and block you.

>68 LolaWalser:
"I think the first step to introducing some clarity in this utter mess of a feature is to unyoke it from the idea of "nationality"."

Yes. We understand that. The CK field "Country (for map)" makes no mention of nationality.

Some authors aren't going to fit neatly into the present-day setup, and some of the complexities will be lost in visualising a catalogue's author nationalities on a present-day world map. Aggregation into a single visualisation will do that.

LolaWalser, you've repeatedly asked this question:

>49 LolaWalser:
It makes me wonder again just what is the point of it. I genuinely don't get it. What is the question this summing up, whether in a list or chart or map, is supposed to answer?

>55 LolaWalser:
I'm just still trying to understand, as above, what question this map is supposed to answer

>61 LolaWalser:
I only asked that trying to understand what question this feature is expected to answer.

>67 LolaWalser:
trying to get a hang on what question you bloomin' think this feature answers

So I'll have a stab.

I think this is the wrong question to ask about this feature, because it is not an "answer" to any one single question, and it wasn't intended as such. Judging it by that standard guarantees a failing grade.

Instead, it can answer many questions LTers might have about their catalogue. I could give you a list, but, frankly, it's late here and I'll just take the short cut. I put it to you that the questions that this feature answers are analogous to the ones answered by the graphs for "Author gender" and "Dead or alive?". Whatever questions those features answer, whatever they are for, those are, mutatis mutandis what this feature is for.

Does that answer your question?

>67 LolaWalser:

What's your solution for representing authors with multiple nationalities, or ambiguous/contested nationalities, and those who were nationals of countries that no longer exist?


Same as it is now. Put some unmappable placeholder in "Country for the map", or pick one - people most familiar with a particular author can make that call on a case-by-case basis. Not worth refusing to have the feature just because it doesn't work for 100% of authors.

>68 LolaWalser:

I think the first step to introducing some clarity in this utter mess of a feature is to unyoke it from the idea of "nationality".

Uh, it already is. There's a "Country for the map" field in CK already, has been for years, which has been mentioned numerous times in this thread.

>71 lorax:

Uh, it already is. There's a "Country for the map" field in CK already, has been for years, which has been mentioned numerous times in this thread.

No, I know, sorry for not clarifying--I meant a field used uniquely for this feature, labelled in some way that indicates a much wider set of values than "nationality". A single category from which the data would be sourced.

I think you can see that as things stand now the messiness is inbuilt in the system--what is getting aggregated means any dozen different things (relating to which I was wondering what question those who want the map see answered by it >49 LolaWalser:), and the output itself is totally ambiguous.

What's the maximum "correct" information such a map conveys? (I'm thinking out loud, not being rhetorical.) As far as I can see, it's some vague geographical association and nothing else. For any given author, from such a map alone it would be impossible to know what the association means--were they born there, sojourned there, been a citizen of some country once and/or currently existing there or what.

You could not, from such a map alone, claim that it tells you what, say, was a given author's nationality/ies, or ethnicity, countries of residence etc. For any given author the represented association may or may not answer any or none of the questions I imagined in >49 LolaWalser:.

That's seriously vague.

Not worth refusing to have the feature just because it doesn't work for 100% of authors.

First, to be sure, and I'm not refusing anything on anyone's behalf--if this could be rejigged conceptually and semantically at least, I wouldn't mind checking it out myself. My concern is, given the current state of this feature and the direction it seems to be driven, with what people will think they are taking away from the map, what it appears or will appear to say.

Put some unmappable placeholder in "Country for the map",

OK, to make sure--this would send the given author into "Other", or maybe exclude them completely? As I said above, to me either would be the better alternative, assuming the current concept, sourcing and labelling can't/won't be changed.

or pick one - people most familiar with a particular author can make that call on a case-by-case basis.

Yeah, well, there's the rub, often there is no good call to make. There are hundreds, thousands of important or popular authors who can't be reduced to single-country association without distorting all kind of data. Naturally there's no way to make it perfect, but at least the distortion could be reduced.

Okay. Clearly there are serious and irreconcilable differences to having this data presented in any aggregate form, and if there's one thing we know from LT it's that if any one person objects, the feature will not be built.

That said, LolaWalser, do you have any objections to either "Nationality" in its currently existing, multi-valued form, or "Country for the Map" in its currently existing single-valued form being exposed somewhere other than on the author pages? Either as a catalog column or as a list of nationalities/countries by author or authors by country/nationality (in which case a single author would appear for all listed nationalities/countries), so that those of us who are less bothered by the ontological problems of what it means to be "from" somewhere and just want to know if we've read any books by authors from New Zealand can answer that question? Because honestly, that gets me 90% of what I want, the visualization was just for pretty.

>73 lorax:

Clearly there are serious and irreconcilable differences to having this data presented in any aggregate form,

This is appallingly unfair from you, at least as far as it's regarding what I wrote in this thread. Nowhere did I object to ANY aggregate form; I pointed out the quite obvious problems with this data and actually made suggestions how the aggregate representation, even from the current data source, might be improved! Down to simply "change the label to change the thinking" and "at least ignore what is ambiguous".

and if there's one thing we know from LT it's that if any one person objects, the feature will not be built.

I don't know if this is true, but I do know it's bound to scapegoat me. Sorry for daring to have opinions on this. I keep forgetting I'm not LT-mainstream enough to have anything worth contributing. Tim, or whoever--please note that I wish to be disregarded on this topic. I won't delete my posts because I hate ruining records.

Lorax, I've always seen you as a super-fastidious and precise person, a data and logic wizard. If the things I pointed out so far don't sincerely strike you as worth considering, as arguments having merit, I'd rather not waste your time further.

I'll answer your question because you asked and I don't want to leave it hanging, not because I think my opinion here is of any importance:

do you have any objections to either "Nationality" in its currently existing, multi-valued form, or "Country for the Map" in its currently existing single-valued form being exposed somewhere other than on the author pages?

Absolutely not, regarding the first part; regarding the second, I would hope there would be some indication of potential limitations of single-value. But generally also no, no objection.

Because honestly, that gets me 90% of what I want, the visualization was just for pretty.

And I hope you get this feature complete with the map, the wait's been long enough. Goodbye.

So this thread has been very active

As at lorax says "...those of us who are less bothered by the ontological problems of what it means to be "from" somewhere and just want to know if we've read any books by authors from New Zealand..."

That's all I want! (for all "country for the map" obviously, but as a NZer I want to know the answer to this one). The data is there isn't it? And we're just wanting it to be exposed, as it was at one stage on the pie chart, but a list, a map something showing the data that is in CK for country for the map field. Did this renewed discussion get us any nearer to this?

wifilibrarian (#75):

The data is there isn't it? And we're just wanting it to be exposed, as it was at one stage on the pie chart, but a list, a map something showing the data that is in CK for country for the map field. Did this renewed discussion get us any nearer to this?

Nope. Still no staff attention.

I don't need it on the map. The old situation - pie chart plus list - was good enough. (For me).

Nicole_VanK (#77)

The old situation - pie chart plus list - was good enough. (For me).

For me, too, and it's a more graceful way of handling countries that don't currently exist. Once I've got the list I can easily generate the map if I want to.

Bump. Another anthology entered in my short fiction account, another look at the bleak pie chart of US/UK/Australia/Canada/Other where "Other" is the second-largest wedge but provides zero information.

There has been talk of making available information that is already there in the database. This RSI, I think, is a prime candidate for that.

Bump.

Can you link to the discussion, please? I'll admit I've started skimming some of the "redesign" threads in the past week, since they're mostly people explaining their own usage in great detail, and I haven't seen much in the way of new thoughts from staff.

>81 lorax:
Hmmm. I could have sworn one of the staff posts in the initial thoughts thread mentioned something about visualisations and graphs. I might have been wishfully misremembering. Dammit!

>81 lorax: and >82 Petroglyph: In the thread that >82 Petroglyph: mentions, nationalities are discussed in posts #69 and #75.
I agree it would be great to make better use of that data.

I finally got around to making a writeup about how to make your own map. Once you have a country for each author (which is quite time-consuming, since you need to search each author individually), it's pretty straightforward.

Instructions:

https://wiki.librarything.com/index.php/User:Lorax/Author_maps

>84 lorax:
That is a lot of manual work! Automating all of that would be nice...

>84 lorax: I already have all of my authors and nationalities in an Excel sheet so I could make a pie chart. I didn't know maps were an option! Thank you so much for this!!!!!!

>84 lorax: citation from your wiki page: "Which option you prefer may depend on which of the Mercator projection and Microsoft you dislike more intensely. I keep going back and forth on that one. " made my day :-)

>85 Petroglyph: I think you can probably script this. More than one author sharing a name might be a problem, so some hybrid with some of the information coming from a wiki page or a spreadsheet and some of it from screenscraping might be needed. The export formats don't really give you an author identity, just the name and that is often ambiguous. :-(

Note to self: just found an example in my own library. Author Ian Stewart links to https://www.librarything.com/author/stewartian-1 which is impossible to guess from the export file, that just contains "Stewart, Ian".
So a script would need to go through each workid and fetch the author link (maybe also Other Authors if you want to include them in the map). For each author link the script will then need to fetch the nationality/country.
It's certainly doable, but not pretty.

>84 lorax:

Re. projections. If you don't like Mercator or Microsoft, is there an free online globe type thing you can use, like that at https://earth.nullschool.net ?

Other projections are available on the site, that'll probably leave you thinking more kindly of Mercator and Microsoft: earth > Projection.

>88 Cynfelyn: Nice, but we're looking for something that can take a color map and apply to individual countries. You could probably go backwards from Mercator and do a custom projection after that. I.e. using Excel to put the colors on the map and then rewrap the map. Only problem is that it will probably look like crap :-)

Ah, I was looking for oocalc maps (because we were talking about Excel) but actually R is the better choice, I think.

I'll take a look at this:
https://www.r-spatial.org/r/2018/10/25/ggplot2-sf.html

Ah, this takes ages to install. Project postponed for some rainy day, methinks. I run into things like "package ‘cowplot’ is not available (for R version 3.4.4) ", blah, blah, blah.

Cynfelyn (#88):

I must be missing something, because I don't see where that site can do the "map numerical values and country names to colors on a map" part which is more or less the entire point of this exercise. Map sites are a dime a dozen - visualization tools to associate values with maps are what we're looking for here.

bnielsen (#89):

Your word of power, for googling in this context, is "choropleth". I know I've done them in R but it was eons ago.

>91 lorax: I also think the word would look fine on a t-shirt :-) BTW I remember doing it in Mathematica once upon a time:

<< Miscellaneous`WorldPlot`
Export"math.pbm", WorldPlot{"Denmark"}, WorldToGraphics -> True
WorldData"Denmark"
{{{3294, 520}, {3395, 488}, {3423, 597}, {3409, 495},
{3464, 633}, {3289, 567}, {3295, 520}}}

As you can see, Mathematica's model of Denmark was a polygon with seven edges, so I stopped experimenting!

ETA1: Just spend some part of a rainy day on this link:
https://www.r-bloggers.com/installation-of-r-3-5-on-ubuntu-18-04-lts-and-tips-fo...

Quote: "Thanks to Dirk Eddelbuettel and Michael Rutter, installing R on Ubuntu is a child’s play !"
Yeah right. Alas, I'm not a child anymore, so it ended in grief.

Note to self: Another nice T-shirt: https://topatoco.com/collections/oglaf/products/og-upright

ETA2:

Mathematica revisited since I have a couple of Raspberry Pi's and Raspbian comes with a free version of Mathematica (Thanks to Stephen Wolfram for that!). I can make a version of a world map with each country colored with its flag, so now I "just" need to take the colors from something I provide. That should be possible in another rainy day or two.

And a link to Lorax's guide to creating Author maps (three cheers for Captain Lorax)
https://wiki.librarything.com/index.php/User:Lorax/Author_maps

So to summarize:
I want to script getting the Nationality for primary author for each book in my library
(Here's a choice to count books or count authors, but leave that aside for now)
Map these country names to a canonical list used in Excel / Google Sheets / Mathematica
Map the numbers to a color map
Produce a map.

Example: book_id, title, work_id
58440641,Professor Stewart's cabinet of mathematical curiosities,6382314

https://www.librarything.com/work/6382314/ provides a link to the author:

https://www.librarything.com/author/stewartian-1

Nationality: UK

Any of the (work_id, author) and (author, nationality) or (author, country_for_map) relations can change, so caching them locally is not a very good idea. On the other hand they probably don't change often, so maybe.

I then need to map the nationalies/country_for_map names to something usable in Excel / Google Sheets / Mathematica. And then find a usable color map and produce a map.

ETA 3: I've not given up yet and R is just out in version 4.0.0 and here's a recipe for something that looks useful.
https://www.r-graph-gallery.com/327-chloropleth-map-from-geojson-with-ggplot2.ht...

Bump.

Thanks for the link. Yes, I wish to know at a glance if I've read any books by authors from New Zealand. It would be incredibly useful.

The stats/memes page is my favorite part of librarything and I wish it was expanded.

It has come to my attention that Tim somehow doesn't know LT can't already do this. Which fills me with absolute great confidence about (a) the monitoring of threads on RSI and (b) the level of knowledge among the staff of the basic functioning of the site. So, bump.

Bump - this feature would make Librarything much more useful to me.

>94 Settings: Thanks for that example. I was thinking of plotting "original language" instead of Nationality but that breaks for Australia, New Zealand and US where original language is just noted as "English".

bnielsen (#97):

I was thinking of plotting "original language" instead of Nationality but that breaks for Australia, New Zealand and US where original language is just noted as "English".

It breaks for most of the world outside of Europe, actually. Does Portuguese mean Portugal or Brazil? Does Spanish mean Spain, Mexico, Chile, Cuba? Does French mean France, Côte d'Ivoire, or parts of Canada? Etc.

Breaks for much of Europe too - Swedish could be Sweden or Finland, Russian could be Russia or Belarus, German could be a bunch of countries. Not to speak of Latin.

And Murakami writes in English but is Japanese. Russian could be Soviet Union or Russia. Yeah, far from perfect. But might do for my books, since most of them are probably Danish. But there's lots and lots of incomplete data. But I wonder how bad it'll turn out to be.

Esperanto, too. In the great diaspora of 1974 the intrepid tailors of Esperan carried their native language beyond the borders of the tiny country throughout the world. Shakespeare was so taken with their language that he wrote a recently-discovered sonnet in it as you probably know.

Bump.

It would be really nice to see this implemented given how many people have decided to "read the world" given the lack of travel in the past year. Building out this tool would really help with that.

Progress!

https://www.librarything.com/topic/144601#7577258

Yay!

Six and a half years later, we can finally mark this one as done. Go check out your "Charts and Graphs" section!

Thanks to everyone here who helped me beat the drum to keep it on the radar, however far-off it was for all those years.

Damn, it feels good to finally have this itch scratched.

I'll miss you all, fellow gadflies. See you in another thread, where we may bug The Powers That Be once more.

>106 lorax: You are a winner! Thank you for keeping this going for so long.

thank you >106 lorax: