Thanks! I was able to get these queries running, making them in a
batch of 100, putting a 5 seconds sleep between them.

Now I've downloaded population data and mixed them so I think I've solved this.

Just to clarify:
1. I need to run both the "simple" and the "latest" query and mix the results.
2. I see that for 95% the cases the "latest" query works and returns
the latest data.
3. 5% of the cases, only the "simple" query works, for example:
https://www.wikidata.org/wiki/Q128323

Is that correct like this?

Zsolt

On Thu, 19 Mar 2020 at 11:10, Benno Fünfstück
<benno.fuenfstueck@mailbox.tu-dresden.de> wrote:

 You can just insert all those items into the `VALUES` statement at the
 top of the query.

 Here is a query to only selects data where the "point in time" qualifier
 is present, and then only gives you the latest version: https://w.wiki/KkM

 Note that this query won't return any results for items where "point in
 time" is not specified on any statement. It's unclear to me how that
 case should be dealt with, from a semantic point of view: what can we do
 if we don't know from which time the data is?

 Regards,
 Benno

 On 18.03.20 23:40, Zsolt Ero wrote:
 Thanks! There is about 5000 item ids in the Natural Earth dataset,
 what would be the best way to get them? Also, how can I get the latest
 data? For example in your query Italy shows 2016, and there is 2017
 and 2020 in there.

 On Wed, 18 Mar 2020 at 23:34, Lucas Werkmeister
 <mail@lucaswerkmeister.de> wrote:
 If you have the item IDs already, the query is relatively simple:

 SELECT ?item ?population WHERE {
   VALUES ?item { wd:Q38 wd:Q148 wd:Q884 }
   ?item wdt:P1082 ?population.
 }

 https://w.wiki/KjA

 You can add more values for the ?item (and spread them across several
 lines as well), the three above are just an example.

 Cheers,
 Lucas

 On 18.03.20 22:58, Zsolt Ero wrote:
 Hi,

 I'm contributing to develop an open source scraper for COVID-19 data
 and we are looking to download the population data from Wikidata for
 regions around the world.

 First, we'd like to get province / state / county items but later on
 probably much finer granularity. We have Wikidata Q id-s from Natural
 Earth, we just don't know how to get the population data from Wikidata
 without scraping. I've seen that there is either a 71 GB gzip JSON
 archive or the query service on https://query.wikidata.org/.

 What I'm looking for would be very simple, just {"Q1234567":
 population} pairs in a JSON, I guess the query service would be ideal,
 but I have no idea how to use it (even after looking at the tutorial).

 Can you help me write this very simple query?

 Zsolt
 Wikidata mailing list
 Wikidata@lists.wikimedia.org
 https://lists.wikimedia.org/mailman/listinfo/wikidata

Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata
Wikidata mailing list
Wikidata@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata