I replied at T183903#4822644 but moving discussion here.

I ran this query, similar to the suggested one:

SELECT CONCAT('https://', project, '.org/wiki/', page_title), SUM(view_count) AS views
FROM wmf.pageview_hourly
WHERE
   year = 2018
   AND agent_type = 'user'
GROUP BY project, page_title
ORDER BY views DESC LIMIT 100;

As suspected it gives us the top 100 most-viewed pages overall, across all projects. The results are still mostly uninteresting since we can't filter by namespace:

https://en.wikipedia.org/wiki/Main_Page    5379645168
https://en.wikipedia.org/wiki/-    876590140
https://en.wikipedia.org/wiki/Special:Search    672831776
https://de.wikipedia.org/wiki/Wikipedia:Hauptseite    461976764
https://it.wikipedia.org/wiki/Pagina_principale    204035882
https://wikidata.org/wiki/Special:ItemDisambiguation    200246468
https://fr.wikipedia.org/wiki/Wikipédia:Accueil_principal    199383354
https://ru.wikipedia.org/wiki/Заглавная_страница    189512927
https://es.wikipedia.org/wiki/Wikipedia:Portada    160300275
https://ja.wikipedia.org/wiki/メインページ    139959749
https://de.wikipedia.org/wiki/Spezial:Suche    110455803
https://es.wikipedia.org/wiki/Especial:Buscar    110299248
https://commons.wikimedia.org/wiki/Special:CreateAccount    103094538
https://fr.wikipedia.org/wiki/Sp?cial:Search    101980574
https://wikidata.org/wiki/Special:BlankPage    89845612
https://en.wikipedia.org/wiki/XHamster    73413875
https://zh.wikipedia.org/wiki/Wikipedia:首页    67342762
https://fr.wikipedia.org/wiki/Spécial:Recherche    58486478
https://es.wikipedia.org/wiki/-    54684470
https://en.wikipedia.org/wiki/Special:CreateAccount    50994694
https://es.wikipedia.org/wiki/Wikipedia:Comunicado_4_julio_2018    49239780
https://nl.wikipedia.org/wiki/Hoofdpagina    46661984
https://commons.wikimedia.org/wiki/Special:Search    46074040
https://ru.wikipedia.org/wiki/Служебная:Поиск    44255332
https://wikidata.org/wiki/-    41799646
https://pl.wikipedia.org/wiki/Wikipedia:Strona_główna    38886611
https://pt.wikipedia.org/wiki/Wikipédia:Página_principal    38098446
https://ar.wikipedia.org/wiki/خاص:بحث    37698677
https://pt.wikipedia.org/wiki/Especial:Pesquisar    36695478
https://en.wikipedia.org/wiki/Deaths_in_2018    36638252
https://it.wikipedia.org/wiki/Wikipedia:Comunicato_3_luglio_2018    35997248
https://ja.wikipedia.org/wiki/特別:検索    35344004
https://en.wikipedia.org/wiki/2018_FIFA_World_Cup    33443484
https://commons.wikimedia.org/wiki/Main_Page    32742629
https://en.wikipedia.org/wiki/Avengers:_Infinity_War    32159499
https://en.wikipedia.org/wiki/Darth_Vader    30968707
https://en.wikipedia.org/wiki/Louis_Tomlinson    30771483
https://de.wikipedia.org/wiki/Special:MyPage/toolserverhelferleinconfig.js    30489158
https://zh.wikipedia.org/wiki/Special:搜索    28566962
https://ja.wikipedia.org/wiki/苦行    28071940
https://ru.wikipedia.org/wiki/Borderlands:_The_Pre-Sequel!    27445616
https://en.wikipedia.org/wiki/Exo_(band)    27017620
https://en.wikipedia.org/wiki/Special:Watchlist    26907252
https://ar.wikipedia.org/wiki/الصفحة_الرئيسية    26882040
https://fr.wikipedia.org/wiki/-    24786718
https://en.wikipedia.org/wiki/Special:ElectronPdf    24276756
https://de.wikipedia.org/wiki/-    23332340
https://kk.wikipedia.org/wiki/Template:AdvancedSiteNotices/ajax    22927585
https://cs.wikipedia.org/wiki/Hlavní_strana    22906759
https://ru.wikipedia.org/wiki/-    22162828
https://it.wikipedia.org/wiki/Speciale:Ricerca    21827111
https://ar.wikipedia.org/wiki/-    21586069
https://en.wikipedia.org/wiki/Black_Panther_(film)    20966096
https://en.wikipedia.org/wiki/Freddie_Mercury    20592463
https://sv.wikipedia.org/wiki/Portal:Huvudsida    20452779
https://en.wikipedia.org/wiki/Special:RecentChangesLinked    20105863
https://en.wikipedia.org/wiki/Elizabeth_II    19237429
https://it.wikipedia.org/wiki/TG_Norba_24    18935078
https://en.wikipedia.org/wiki/List_of_United_States_stand-up_comedians    18462069
https://en.wikipedia.org/wiki/List_of_stand-up_comedians    18453179
https://en.wikipedia.org/wiki/List_of_Australian_stand-up_comedians    18391830
https://en.wikipedia.org/wiki/List_of_Canadian_stand-up_comedians    18383631
https://en.wikipedia.org/wiki/List_of_British_stand-up_comedians    18379310
https://en.wikipedia.org/wiki/Stephen_Hawking    18236502
https://zh.wikipedia.org/wiki/-    18100093
https://en.wikipedia.org/wiki/Meghan_Markle    18029873
https://ja.wikipedia.org/wiki/-    17640804
https://en.wikipedia.org/wiki/List_of_Marvel_Cinematic_Universe_films    17580795
https://en.wikipedia.org/wiki/Cristiano_Ronaldo    17488886 (edited) 
https://fi.wikipedia.org/wiki/Wikipedia:Etusivu    17160316
https://en.wikipedia.org/wiki/Cardi_B    17010723
https://en.wikipedia.org/wiki/Elon_Musk    16899563
https://en.wikipedia.org/wiki/Donald_Trump    16729532
https://en.wiktionary.org/wiki/Wiktionary:Main_Page    15981543
https://en.wikipedia.org/wiki/BTS_(band)    15323915
https://en.wikipedia.org/wiki/Portal:Current_events    15239781
https://pt.wikipedia.org/wiki/-    15125720
https://de.wikipedia.org/wiki/Special:Search    14998199
https://pl.wikipedia.org/wiki/Specjalna:Szukaj    14981937
https://en.wikipedia.org/wiki/XXXTentacion    14735395
https://en.wikipedia.org/wiki/United_States    14261908
https://en.wikipedia.org/wiki/List_of_Bollywood_films_of_2018    14000321
https://nl.wikipedia.org/wiki/Speciaal:Zoeken    13707855
https://en.wikipedia.org/wiki/Special:CiteThisPage    13465627
https://en.wiktionary.org/wiki/Special:Search    13387173
https://ko.wikipedia.org/wiki/위키백과:대문    13175873
https://en.wikipedia.org/wiki/Special:MobileMenu    13172982
https://en.wikipedia.org/wiki/Tunnel    12791300
https://id.wikipedia.org/wiki/Istimewa:Pencarian    12696208
https://en.wikipedia.org/wiki/List_of_highest-grossing_films    12289384
https://en.wikipedia.org/wiki/XXX_(franchise)    12198573
https://commons.wikimedia.org/wiki/Special:RecentChangesLinked    12005056
https://cs.wikipedia.org/wiki/Speciální:Hledání    12003088
https://en.wikipedia.org/wiki/LeBron_James    11975345
https://en.wikipedia.org/wiki/File:ITVBe_logo_2014-.svg    11909554
https://it.wikipedia.org/wiki/-    11625413
https://en.wikipedia.org/wiki/George_H._W._Bush    11603005
https://de.wikipedia.org/wiki/Spezial:Beobachtungsliste    11591257
https://en.wikipedia.org/wiki/Anthony_Bourdain    11579001
https://en.wikipedia.org/wiki/6ix9ine    11575908

This took about 45 minutes to complete.

I'm not the best at SQL, but borrowing from XTools Top Edits -- which shows the top 100 most-edited pages by a user for each individual namespace -- I think we might need a more complicated query like:

SELECT c.project, c.page, c.views 
FROM
(
  SELECT b.project, b.page, b.views,
    @rn := IF(@project = b.project, @rn + 1, 1) AS row_number,
    @project := b.project AS dummy
  FROM (
    SELECT project,
      CONCAT('https://', project, '.org/wiki/', page_title) AS page,
      SUM(view_count) AS views
    FROM wmf.pageview_hourly
      WHERE year = 2018
        AND agent_type = 'user'
      GROUP BY project, page_title
    ) AS b
    JOIN (SELECT @project := NULL, @rn := 0) AS vars
    ORDER BY b.project ASC, b.views DESC
  ) AS c
WHERE c.row_number <= 100

However here Hive doesn't seem to like the variable notation @rn := ..., etc.

Additionally, while it will slow down the query even more, we might add some logic like page_title NOT LIKE "%:%" to try to filter out non-mainspace, though of course there could be page titles with colons in them that are in the mainspace.

It might be easier and/or more feasible to just query for each project individually, perhaps the top 10 Wikipedias? Then whatever else, as requested.

In T211827#4822756, @MusikAnimal wrote:

I replied at T183903#4822644 but moving discussion here.

I ran this query, similar to the suggested one:

...

As suspected it gives us the top 100 most-viewed pages overall, across all projects. The results are still mostly uninteresting since we can't filter by namespace:

Actually, a namespace_id field was added to the pageview_hourly during 2017 (T156993), so now is the first time we can try to use it for this purpose. While it is not yet 100% accurate in the sense that there remains a small amount of pageviews that still don't have value set for this field (mainly from the mobile apps IIRC), it should work for the present purpose of excluding most non-mainspace pages from the top 100 or so pages.

In T211827#4822761, @Tbayer wrote:

Actually, a namespace_id field was added to the pageview_hourly during 2017 (T156993), so now is the first time we can try to use it for this purpose. While it is not yet 100% accurate in the sense that there remains a small amount of pageviews that still don't have value set for this field (mainly from the mobile apps IIRC), it should work for the present purpose of excluding most non-mainspace pages from the top 100 or so pages.

That's fantastic! I ran your query from last year, adding in a namespace check, and this is what I got for enwiki:

As you know, some of these such as XHamster are widely viewed as false positives, at least in terms of authentic public interest.

The query strangely took around the same amount of time as the one that scanned all projects.

Some ideas (for @Quiddity mostly)

• Given the disparity of views among enwiki and other wikis a top 100 over the year will likely (as you might have seen) only give you results for enwiki.

• You can approximate this results with api as the top 100 for the year can be computed probably from the top 1000 per every month. Some redaction would be needed as you need to remove main page , etc. An example request that returns a 1000 records for the month: https://wikimedia.org/api/rest_v1/metrics/pageviews/top/en.wikipedia.org/all-access/2018/11/all-days

• This data is still affected by bots and non-intentional traffic, the best example of which might be XHamster page.

Commented with a better query fixing my naive first attempt here: T183903#4824489

@Nuria Yes, sorry that my description was ambiguous! We want separate lists for each Wikipedia. I.e. various Chapters want language-specific lists, to share/answer with their own press contacts.

In T211827#4822756, @MusikAnimal wrote:

It might be easier and/or more feasible to just query for each project individually, perhaps the top 10 Wikipedias? Then whatever else, as requested.

Yes, Top 10, or maybe Top ~40 (? would that cover most of the 37 Chapters?) would be good to start with.

p.s. Enwiki is basically already covered, by Andrew West who does it every year, and started the version for 2018 a few days ago. https://en.wikipedia.org/wiki/User:West.andrew.g/2018_Popular_pages

I'm happy to help run the query on some of the top projects if we can't find a better way. @Quiddity Do you plan on publishing this anywhere public, such as the wiki? Meta or the like? I'd love to link to the results from https://tools.wmflabs.org/topviews

I do think we should wait till January 1 (perhaps the 2nd) before going all out. If a major event happens between now and then, the corresponding articles could skyrocket their way to the top.

In T211827#4824638, @MusikAnimal wrote:

@Quiddity Do you plan on [...]

I was originally planning on spending 2 minutes writing a reply to an email, just pointing everyone towards Topviews as a solution for all their needs! But then I found it doesn't include yearly yet, so I fell down the rabbithole of investigating the things that lead towards T154381. (many links... much reading... hours later...). I ended up writing this task to save anyone else the bother of going through all that again! I'll CC @EdErhart-WMF who might be able/willing to do more of the actual decision-making and organising what people need and where they want it placed. /me runs away.

Hello everyone! We've been talking on the Communications Committee's mailing list about this. @MusikAnimal, would it be possible to get a list of the most-viewed articles in the following languages shortly after the New Year?

German, Hebrew, Polish, Swedish, and Chinese.

Thank you!

In T211827#4838648, @EdErhart-WMF wrote:

Hello everyone! We've been talking on the Communications Committee's mailing list about this. @MusikAnimal, would it be possible to get a list of the most-viewed articles in the following languages shortly after the New Year?

German, Hebrew, Polish, Swedish, and Chinese.

Thank you!

Yes, I'm happy to help :) Five wikis is not that much.

I spoke with @Nuria about this briefly over IRC -- I have a list of false positives from crowd-sourced data provided by users of https://tools.wmflabs.org/topviews, organized by project, platform and date. I have listed this data below should it interest anyone, though much of it does not affect the request projects. For most of these I confirmed as false positives by querying Hive and seeing if 90+% of traffic came from a single city, where that city isn't particularly large. Some FPs were ascertained solely because of a crazy unlikely % mobile/desktop ratio. That said, there's no guarantee this is accurate, but I'm fairly confident these are FPs. null means all possible values for that column:

+1 to exclude those on list and and @EdErhart-WMF to include the fact that exclusions come from crowdsourced data from our community. I think the excluded titles make up for a blogposts themselves.

If you can add Greek, we'd be very happy :-)

I just made a query that yields the result for all wikis simultaneously (P7945, based on @Milimetric's earlier draft) and also includes some helpful additions like the percentage of mobile views for each page (for easier anomaly detection). See T183903#4847361 for details - perhaps @MusikAnimal might be interested in running this too and posting the result in suitable form for all interested parties. The result will still need some manual vetting for each wiki.

CCing @Reke and @Shangkuanlc who have an interest in this topic as well.

In T211827#4847368, @Tbayer wrote:

I just made a query that yields the result for all wikis simultaneously (P7945, based on milimetric's earlier draft) and also includes some helpful additions like the percentage of mobile views for each page (for easier anomaly detection). See T183903#4847361 for details - perhaps MusikAnimal might be interested in running this too and posting the result in suitable form for all interested parties. The result will still need some manual vetting for each wiki.

Going to reply here. The query is amazing!! Thank you and @Milimetric :) I ran it myself and indeed it took maybe 45 minutes. Unfortunately I can't figure how to export to disk, and the number of lines returned in the result set exceeds the max number of lines in my terminal. So, I can't see the full results. How did you do it?

I'm happy to help filter out the anomalies. I will point out that the mobile/desktop viewership % doesn't seem to work on all projects, e.g. those where mobile devices aren't as popular. Hopefully we can tell by just looking at the average % for the full list on that wiki. 800+ wikis is an awful lot, so maybe we could split up the work, or just focus on the large wikis?

Using a fancy text editor, I think wikifying the list will be pretty straightforward. Should we just dump the results on Meta, one page per wiki? Probably will limit this to the wikis of interest, to avoid having to get bot approval (800+ pages is a lot!).

@EdErhart-WMF @geraki Do you need any other formats? CSV or HTML with some specific structure?

In T211827#4847998, @MusikAnimal wrote:

In T211827#4847368, @Tbayer wrote:

I just made a query that yields the result for all wikis simultaneously (P7945, based on milimetric's earlier draft) and also includes some helpful additions like the percentage of mobile views for each page (for easier anomaly detection). See T183903#4847361 for details - perhaps MusikAnimal might be interested in running this too and posting the result in suitable form for all interested parties. The result will still need some manual vetting for each wiki.

Going to reply here. The query is amazing!! Thank you and @Milimetric :) I ran it myself and indeed it took maybe 45 minutes. Unfortunately I can't figure how to export to disk, and the number of lines returned in the result set exceeds the max number of lines in my terminal. So, I can't see the full results. How did you do it?

You can use Hive in batch mode and save the result in a file, see e.g.
https://wikitech.wikimedia.org/wiki/Analytics/Systems/Cluster/Hive/Queries#Batch_mode .
Personally I often use this from within a shell script that also sends me an email with the result as soon as the query has completed: P7948
(For large result files such as this, SCP may be a better option though than email for the purpose of transfering the file from the analytics server to your own machine.)

I'm happy to help filter out the anomalies. I will point out that the mobile/desktop viewership % doesn't seem to work on all projects, e.g. those where mobile devices aren't as popular. Hopefully we can tell by just looking at the average % for the full list on that wiki. 800+ wikis is an awful lot, so maybe we could split up the work, or just focus on the large wikis?

Personally I think it would be fine to just publish the raw results with the known general caveats, and leave it to interested community members or affiliate organizations to vet them and compile them into a list for an individual wiki. (That is also a task that could benefit from knowledge of the corresponding language and culture.)

Using a fancy text editor, I think wikifying the list will be pretty straightforward. Should we just dump the results on Meta, one page per wiki? Probably will limit this to the wikis of interest, to avoid having to get bot approval (800+ pages is a lot!).

Would the TSV fit into a Google spreadsheet?

@EdErhart-WMF @geraki Do you need any other formats? CSV or HTML with some specific structure?

Yes, TSV would be great and can work with both Excel and Google spreadsheet.

Would the TSV fit into a Google spreadsheet?

@EdErhart-WMF @geraki Do you need any other formats? CSV or HTML with some specific structure?

It would be great if you can provide a list for ms.wikipedia :D

It would be great if in 11 months anyone could create the lists that they want themselves, by following some documentation.

The file produced by Tbayer's query is too big to upload to Phabricator, apparently, so I've attached each requested wiki below. All credit goes to Tbayer and Milimetric for collecting this data! I just ran the query they wrote :)

I tested it and it did work in Google Sheets. Per above I did not vet the data myself. There are some known data anomalies listed at T211827#4838895 that you can refer to, or go by the percentage of mobile traffic (last column in the TSV), at your discretion.

I'm going to try to get this data integrated into https://tools.wmflabs.org/topviews somehow (maybe even the top 1000 pages), unless there's a possibility T154381 will get addressed soon? If I manage to get it to work, I'll do the same for 2017 and 2016.

Hi @MusikAnimal, could you export data for Serbian Wikipedia (srwiki) as well?

In T211827#4852653, @dungodung wrote:

Hi @MusikAnimal, could you export data for Serbian Wikipedia (srwiki) as well?

I did manage to come up with a system to show the yearly data (for any project) in https://tools.wmflabs.org/topviews. It is still in development but hopefully will have it deployed tomorrow :)

@MusikAnimal: Yay! Feel free to claim T154446

I'm going to try to get this data integrated into https://tools.wmflabs.org/topviews somehow (maybe even the top 1000 pages), unless there's a possibility T154381 will get addressed soon? If I manage to get it to work, I'll do the same for 2017 and 2016.

Thank you @MusikAnimal for creating the tsv file. I have listed the [[ https://docs.google.com/spreadsheets/d/1xMfNcMFJibJkLFEscpRxuFiNIdNZgeDaWJksY-2jASw/edit?usp=sharing

In T211827#4852637, @MusikAnimal wrote:

The file produced by Tbayer's query is too big to upload to Phabricator, apparently, so I've attached each requested wiki below. All credit goes to Tbayer and Milimetric for collecting this data! I just ran the query they wrote :)

@MusikAnimal Could you please run the querry for the Persian Wikipedia (fawiki) as well?

In T211827#4854421, @4nn1l2 wrote:

MusikAnimal Could you please run the querry for the Persian Wikipedia (fawiki) as well?

In T211827#4854404, @Shangkuanlc wrote:

I have listed the Chinese Wikipedia articles by the columns you indicated. However, I am confused by the above quote. What is the typical anomalies that you suggest? I have highlight in yellow in the google spreadsheet for the links that receives the pageviews by mobile traffic under 20 %. Is that the only criteria that I should tick out?

It depends on the wiki. On English Wikipedia, it's unusual for there to be very low mobile viewership, or even very low desktop. I would look at what seems normal on your wiki (based on most of the entries in your most-viewed report), and judge from that. Other tactics commonly used:

• Does it make sense that the article would be in the most-viewed? Generally I would expect topical subjects (e.g. were in the news)
• When looking at per-day pageviews, do you see patterns that look bot-like? Usually for genuine traffic, I see a spike, then traffic slowly gets smaller over time as public interest declines. If you see a huge spike of pageviews in a single day or short time period, then suddenly it goes back to lower pageviews, that might be suspect (unless it was "today's featured article" or the like)
• How many edits were made in the period? Generally high viewership will also attract editing. It may be tedious to get this data by hand, but hopefully later today I'll have the yearly data on https://tools.wmflabs.org/topviews so you can quickly see what editing activity was like for each article.

From the crowd-sourced data I have three Chinese articles listed as likely false positives:

• 郭建宏 (製作人) -- Almost all traffic is from desktop. Pageviews chart suggests quick, isolated periods of interest. Compare with 如懿传 for example where you can see traffic slowly tapered off, which to me seems more natural.
• 維基媒體基金會 -- This is a redirect that had zero edits in 2018. The pageviews chart looks suspicious, especially from April through August where there were almost exactly the same number of pageviews every day.
• Wiki -- The pageviews chart looks similar to 維基媒體基金會. Only 7 edits for the year.

If there are any other articles you're unsure about, we can look at private data (such as where traffic originated from geographically), but this is tedious.

For the record, to others with access to the analytics cluster, this is an example query I use to confirm anomalies (or "false positives" as I call them):

SELECT city, SUM(view_count) AS viewcount
FROM pageview_hourly
WHERE page_title = "Foo bar"
 AND project = "en.wikipedia"
 AND year = 2018
 AND month = 11
 AND day = 11
GROUP BY city
ORDER BY viewcount DESC
LIMIT 5;

Usually a very high amount of traffic will come from a single random city, and that city is often is not very large, or in a country that doesn't have many speakers of the language the article was written in.

Topviews now has yearly data for 2017 and 2018: https://tools.wmflabs.org/topviews/?project=zh.wikipedia.org&date=2018 and I'll be adding 2016 shortly.

Due to storage concerns, I've only included the top 500 pages in the mainspace, and only for 'all-access' platform (desktop and mobile combined). I admittedly cobbled together the code as quick as possible. There may be some UI bugs (e.g. you click on thing and another thing doesn't work anymore), but the data you see should be accurate.

Many of the false positives mentioned above (T211827#4855654 specifically) have been removed from Topviews. I'm going through other large projects and removing other obvious false positives. You can help, too! See https://tools.wmflabs.org/topviews/faq/#false_positive for more info. You can also see which pages were removed in the tool itself, look for the link at "[N pages] have been automatically excluded" below the Search bar.

Hope this helps!

@MusikAnimal Can you do for Malay Wikipedia (ms.wikipedia.org)? The one you did was Malayalam (ml.wikipedia.org)

In T211827#4856531, @Taufik wrote:

@MusikAnimal Can you do for Malay Wikipedia (ms.wikipedia.org)? The one you did was Malayalam (ml.wikipedia.org)

Sorry about that. Note you can now get the data at https://tools.wmflabs.org/topviews/?project=ms.wikipedia.org&date=2018. This also includes editing statistics. You can then use Download → CSV to export as CSV. I can add TSV if needed. There is also sort of a bug here -- the CSV won't include editing data except for the articles shown on the screen. So just click "Show more" at the bottom 5 times to load the rest of the editing data.

Hopefully the above will suffice for others, but anyway here's the TSV for ms.wikipedia (does not include editing stats):

In T211827#4856171, @MusikAnimal wrote:

Topviews now has yearly data for 2017 and 2018: https://tools.wmflabs.org/topviews/?project=zh.wikipedia.org&date=2018 and I'll be adding 2016 shortly.

@MusikAnimal Thanks! There is a weird issue with the links, however. I can see the topviews list for the Persian Wikipedia following this link https://tools.wmflabs.org/topviews/?project=fa.wikipedia.org&platform=all-access&date=2018&excludes= on Chrome v. 67, but not on Firefox v. 64! On Firefox, it jumsps to the monthly viewes (December 2018).

在T211827#4855654中，@MusikAnimal写道：

In T211827#4854421, @4nn1l2 wrote:

• 維基媒體基金會 -- This is a redirect that had zero edits in 2018. The pageviews chart looks suspicious, especially from April through August where there were almost exactly the same number of pageviews every day.
• Wiki -- The pageviews chart looks similar to 維基媒體基金會. Only 7 edits for the year.

If there are any other articles you're unsure about, we can look at private data (such as where traffic originated from geographically), but this is tedious.

About this, see T172379

I guess this task can be closed now?

BTW it might be nice to collect some of the pieces have been published based on this data, e.g.:

• https://www.wikimedia.fr/2019/01/05/articles-wikipedia-plus-consultes-2018/
• http://blog.wikimedia.pt/2019/01/artigos-mais-vistos-em-2018/