Template talk:COVID-19 pandemic death rates
![]() | This template does not require a rating on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||
|
Always visible header row as with other template?
[edit]See:
Note that the header row there on that template stays visible as one scrolls. Please add that function to this template here. --Timeshifter (talk) 11:39, 1 September 2020 (UTC)
![]() | This help request has been answered. If you need more help, you can , contact the responding user(s) directly on their user talk page, or consider visiting the Teahouse. |
I do not know how to do this. Can an expert advise please? Dudley Miles (talk) 13:49, 1 September 2020 (UTC)
- Not possible with the current setup. In the other template it's one table, but here you're transcluding a table into another table. Either this template or {{COVID-19 pandemic death rates by country}} will need a significant rewrite to manage it. Primefac (talk) 21:16, 2 September 2020 (UTC)
- Primefac and Dudley Miles. See User:Timeshifter/Sandbox119. The header row stays visible.
- The table is updated by copying the data wikitext (as is) from:
- Template:COVID-19 pandemic death rates by country
- It may be better to not be dependent on that template. By creating another template without an external stylesheet.
- Can this CSS below be added to here?:
- Template:COVID-19 pandemic data/styles.css
/* styling for first column */ .covid19-deathbycountry tr>:first-child { text-align:left; }
- I don't know why the text in the country column is wrapping. It isn't wrapping at the template I used as a model:
- Template:COVID-19 pandemic data.
- --Timeshifter (talk) 15:14, 4 September 2020 (UTC)
- I fixed the country column text wrapping at User:Timeshifter/Sandbox119. --Timeshifter (talk) 00:59, 5 September 2020 (UTC)
CSS to left-align column of country names. When country names are in the first column
[edit]Some CSS needs to be added to here:
This way we don't need to source this CSS:
It has this:
/* styling for first column */ .covid19-deathbycountry tr>:first-child { text-align:left; }
I will change it to this before adding it to the other CSS page:
/* Column of country names left aligned when names are in first column */ /* For John Hopkins table. */ .covid19-countrynames tr>:first-child { text-align:left; }
This is because the other table has the country names in the second column. Its first column is for flags.
Whereas the John Hopkins table has the country names (along with the flags) in the first column.
I changed the CSS selector name to a more descriptive name: covid19-countrynames
I tried it out and it works. --Timeshifter (talk) 06:28, 6 September 2020 (UTC)
- Thanks for dealing with this. CSS is beyond my technical knowledge. Dudley Miles (talk) 09:04, 6 September 2020 (UTC)
Data should be for the date at the top of the table
[edit]Mithila Madawa Gunathilake and others doing single country updates. Table is currently for Nov 19, not Nov 20. If you want to update to Nov 20, then please do the whole table. It only takes a few minutes. See how at doc page: Template:COVID-19 pandemic death rates/doc.
Otherwise the data does not match the source for the date indicated at the top of the table. --Timeshifter (talk) 02:12, 21 November 2020 (UTC)
Mistake in Johns Hopkins data
[edit]I know that since Johns Hopkins is used as a source the data should be used as is, but I noticed a pretty significant mistake. The number of confirmed deaths and reported cases has been confused between Congo and DR Congo, leading to these fields (and the rest of them since they are derived from the amount of deaths) being very wrong, almost by an entire order of magnitude. Should we fix this (and mention that the source has been modified if necessary)? Antondimak (talk) 12:43, 11 May 2021 (UTC)
- Antondimak sorry I have only just noticed this post. I see you are correct that John Hopkins has transposed the figures for the two countries. There are some errors in the JH data and I am reluctant to correct this one. It would take significant additional work daily compared with just copying and could imply that the other figures are all correct. I will add a note. Dudley Miles (talk) 09:18, 11 June 2021 (UTC)
Peru
[edit](Copied from User talk:Dudley Miles). Dudley, someone pointed out that the Peru deaths/100,000 figure on the Covid-19 pandemic by country & territory page was way out. I've patched it, but can you check your population figures for Peru, which I presume was the culprit. Thanks. Chris55 (talk) 08:22, 11 June 2021 (UTC)
- Chris55 Can you clarify how you arrive at the figure of 224.30. Deaths are 187,157 and population according to Peru is 32,824,358. This gives a deaths per 100,000 of 570.20. This is close to the John Hopkins figure, the basis of the table, of 575.68, and easily accounted for as they are obviously using a source with a slightly different population figure. Dudley Miles (talk) 08:54, 11 June 2021 (UTC)
- This table is copied daily from the John Hopkins source. There are some dubious figures in it but I copy the source unamended as I would otherwise be using WP:OR to give an incorrect copy of the source. Dudley Miles (talk) 09:05, 11 June 2021 (UTC)
- You're right, Dudley, sorry about that. I've reverted the edit. I was working from the ECDC data but predating the big Peru revision, so was only assuming 76,095 deaths. I thought I'd updated my spreadsheet but I hadn't. I've now cross-checked with the John Hopkins data and they agree. I've reverted the change. Chris55 (talk) 12:06, 11 June 2021 (UTC)
- This table is copied daily from the John Hopkins source. There are some dubious figures in it but I copy the source unamended as I would otherwise be using WP:OR to give an incorrect copy of the source. Dudley Miles (talk) 09:05, 11 June 2021 (UTC)
Now, the text in the "Peru" column reads "nan". Is there any way to ascertain the number of deaths per 100,000 people in Peru (the highest confirmed death toll per 100,000 people in the world)?--2A02:AB04:236:E600:9DFC:E206:6F62:9314 (talk) 19:11, 20 June 2021 (UTC)
- That is a typo now corrected. Dudley Miles (talk) 20:05, 20 June 2021 (UTC)
Case Fatality Rate Rounding
[edit]The case fatality rates are give to two decimal places, but have clearly been rounded to one d.p. This gives a false impression of accuracy: they should either be given to 1 d.p. (ie removing the 0 on the end of each), or given to two d.p. — Preceding unsigned comment added by 86.179.88.201 (talk) 09:51, 29 July 2021 (UTC)
- A good point. This error occurs during transfer from the Johns Hopkins table, and I will try to correct it in tomorrow's update. Dudley Miles (talk) 20:54, 29 July 2021 (UTC)
Automation of updates
[edit]Hello! I'm Tol, and I've been working on automating updating of COVID-19 data templates using Module:COVID-19 data. I've discussed this with Timeshifter on my talk (permalink), and I'd like to see what the general opinion is. This would probably entail replacing deaths per hundred thousand with deaths per million, and may also include removing the case–fatality ratio (I could write some custom code to calculate it, but I'd prefer not to, and Timeshifter has brought up problems with it depending on testing rates). What do you think? (Also posted at Template talk:COVID-19 pandemic death rates by country#Automation of updates) Tol (talk | contribs) @ 17:48, 23 September 2021 (UTC)
- I have no problem in principle for automating the updates to {{COVID-19 pandemic death rates by country}}, which I've being doing myself since April. I use the python script near the end of the template talk page. I do make a few small changes after running the script: I update the date at the top of the template; I add a manually calculated figure for deaths per 100k for Eritrea, which the source for some reason gives as NaN and the python script converts to a dash, and then move the entry for Eritrea to the correct sort order; and I adjust the entry for China to link to COVID-19 pandemic in mainland China because the link without the word "mainland" goes to a disambiguation page. These minor adjustments take me about an extra minute per day.-gadfium 18:53, 23 September 2021 (UTC)
- And I've just noticed that COVID-19 pandemic in China is no longer a disambiguation page so I can stop making that adjustment. I see that this template {{COVID-19 pandemic death rates}} has the same error for Eritrea which no one has bothered to fix.-gadfium 18:58, 23 September 2021 (UTC)
- @Gadfium: Alright. I've made a sandbox version at Template:COVID-19 pandemic death rates by country/sandbox which implements the module. Does it look good to you? Tol (talk | contribs) @ 19:28, 23 September 2021 (UTC)
- You are probably aware that there are parts of your draft table where the data is "Unknown" or there is "No data". I presume you can suppress these rows, because they're certainly not useful. There are several ambiguous entries, e.g. for Georgia, Ireland and Congo. The existing table handles these properly (my mention of China above was a case where it didn't handle it a couple of months ago). I think the extra column with flags is unnecessary and ugly but that's just my opinion. I realise the existing table has flags but they're better integrated with the country names. Because the title of the template is "Death rates..." the default sort should be by the last column. The death rates differ between the existing table and the draft, although the death figures seem to be the same, so presumably the two sources are using different population figures for each country. I'm not sure whether this matters.-gadfium 20:00, 23 September 2021 (UTC)
- @Gadfium: I just made a fix so rows with no data are skipped, and added a
noflag
parameter which suppresses the flag. How does it look now? Tol (talk | contribs) @ 21:33, 23 September 2021 (UTC)- Much better. There are still some entries at the bottom where countries have had COVID cases but no deaths. You could either replace "Unknown" with "0", or drop these entries entirely. Several of the ambiguous entries disappeared with your change, but Georgia, Ireland and Congo still link to disambiguation pages, e.g. COVID-19 pandemic in Congo — Preceding unsigned comment added by Gadfium (talk • contribs) 21:46, 23 September 2021 (UTC)
- @Gadfium: I just made a fix so rows with no data are skipped, and added a
- You are probably aware that there are parts of your draft table where the data is "Unknown" or there is "No data". I presume you can suppress these rows, because they're certainly not useful. There are several ambiguous entries, e.g. for Georgia, Ireland and Congo. The existing table handles these properly (my mention of China above was a case where it didn't handle it a couple of months ago). I think the extra column with flags is unnecessary and ugly but that's just my opinion. I realise the existing table has flags but they're better integrated with the country names. Because the title of the template is "Death rates..." the default sort should be by the last column. The death rates differ between the existing table and the draft, although the death figures seem to be the same, so presumably the two sources are using different population figures for each country. I'm not sure whether this matters.-gadfium 20:00, 23 September 2021 (UTC)
I suggest putting the columns in this order below. With the table initially sorted via the "Deaths per million" column. I think it is the most significant number indicating how hard a country or subnational area has been hit by Covid.
! scope=col | Country, [[subnational area|area]] ! scope=col | Deaths per million ! scope=col | Deaths ! scope=col | Cases
Many people want the flags, and some are hardcore about it. :)
I don't care either way. As long as the country names are left aligned.
--Timeshifter (talk) 22:00, 23 September 2021 (UTC)
- @Gadfium: I've fixed Ireland and Congo at Template:COVID-19 data/data. Georgia will need a manual override in the module, which I'm working on now. The only (easy) way to get dropping rows to work is to only drop entirely blank rows. If the cases column is removed (as @Timeshifter suggested) then only locations with deaths will show. @Timeshifter: I usually use "location" instead of "country" or "area" because it's neutral and short (think of places like Hong Kong). Implementing custom sorting would also be hard — it would have to be done in the module because there's no way to sort a table on page load yet. Tol (talk | contribs) @ 22:24, 23 September 2021 (UTC)
- Thanks for your work on this. As I update each day I would obviously be happy to be saved the work. I think it is crucial that the table sorts by default on the death rate and that the death rates are copied from the JH table. It should be a copy of the JH table in user friendly form, not having unreferenced changes due to different population figures or for any other reason. Dudley Miles (talk) — Preceding undated comment added 22:30, 23 September 2021 (UTC)
- "Location" works for me. If the module can sort the death rate column that would be great. Can the module rearrange the columns so that the death rate column is the first column? That would make the table more intuitive when people see right away that the table is sorted by the death rate.
- Is this the source of the data?:
- Template:COVID-19 data/Cite
- If so, then it is authoritative and reliable. Our World in Data is considered to be a reliable source. See:
- COVID-19 pandemic by country and territory#Notes
- --Timeshifter (talk) 22:44, 23 September 2021 (UTC)
- @Dudley Miles & @Timeshifter: Our World in Data (OWID) is the source of all data from the module. @Dudley Miles: If you think so, then I'll get to work implementing custom sorting. I'm not sure what "JH" is, but if it's Johns Hopkins University, that's where OWID (and so we) get our data from. @Timeshifter: Custom ordering should be easier than custom sorting. I'll work on that, too. Tol (talk | contribs) @ 22:50, 23 September 2021 (UTC)
- @Dudley Miles & @Timeshifter: Apparently, I'd already implemented custom ordering! I implemented custom sorting (and did some general fixes for better compatibility with sorting and Template:Static row numbers). Tol (talk | contribs) @ 23:19, 23 September 2021 (UTC)
- Looks very good now. If the consensus is to have country flags, please put them in the same column as the country names, as {{COVID-19 pandemic death rates by country}} does. And since you've already put far more effort into this than you were probably expecting to, how much more would it be to add back the case fatality rate?-gadfium 23:26, 23 September 2021 (UTC)
- @Gadfium: For country flags, that'll be difficult. If everyone's fine with leaving them out, then I'd prefer that. Adding the case fatality rate probably wouldn't be hard on-wiki, but it would require a rewrite of the bot task's mapper (maps OWID to our JSON page). I can try to do that too, though. Tol (talk | contribs) @ 23:35, 23 September 2021 (UTC)
- Looks very good now. If the consensus is to have country flags, please put them in the same column as the country names, as {{COVID-19 pandemic death rates by country}} does. And since you've already put far more effort into this than you were probably expecting to, how much more would it be to add back the case fatality rate?-gadfium 23:26, 23 September 2021 (UTC)
- @Dudley Miles & @Timeshifter: Apparently, I'd already implemented custom ordering! I implemented custom sorting (and did some general fixes for better compatibility with sorting and Template:Static row numbers). Tol (talk | contribs) @ 23:19, 23 September 2021 (UTC)
- @Dudley Miles & @Timeshifter: Our World in Data (OWID) is the source of all data from the module. @Dudley Miles: If you think so, then I'll get to work implementing custom sorting. I'm not sure what "JH" is, but if it's Johns Hopkins University, that's where OWID (and so we) get our data from. @Timeshifter: Custom ordering should be easier than custom sorting. I'll work on that, too. Tol (talk | contribs) @ 22:50, 23 September 2021 (UTC)
Looks great. I now think we should keep the cases column. It looks like some of the small islands may be claiming zero deaths from Covid. So cases is all we have for them. As for flags we probably should keep them to avoid problems. As for case fatality rate I think that if we do calculations, then it is approaching original research. If we are just copying that rate from OWID, then I guess it can be used here. I don't like the case fatality rate, but I will go along with whatever decision you all make. --Timeshifter (talk) 23:35, 23 September 2021 (UTC)
- @Timeshifter: I also dislike the case fatality rate (it only really works with a better measure of cases such as excess deaths). I don't understand what problems removing the flag would cause — could you elaborate? Tol (talk | contribs) @ 23:44, 23 September 2021 (UTC)
- I almost guarantee that someone will complain. But if it is difficult to include the flags, then don't bother. Since it is a complicated bot template, and not a simple table, then people will understand. --Timeshifter (talk) 00:01, 24 September 2021 (UTC)
- The reference for the table has a note "This table copies figures unamended from a JHU (Johns Hopkins University) table". I do not think it matters whether the flags and case fatality rate are retained, but I think it is crucial that the figures are the same - as the note says - and that the default sort is the death rate. If I understand correctly you have already implemented the default sort but I am not sure about the identical figures. Dudley Miles (talk) 08:17, 24 September 2021 (UTC)
- Dudley Miles. Tol has this reference for the sandbox table:
- https://ourworldindata.org/coronavirus
- It has this link:
- "Download our complete dataset of COVID-19 metrics on GitHub. It’s open access and free for anyone to use.".
- That is the source data for all OWID maps, tables, and graphs. --Timeshifter (talk) 10:55, 24 September 2021 (UTC)
- Timeshifter. Well that is obviously fine if the table is going to use another reliable source and I can stop updating and leave it to more expert editors. Dudley Miles (talk) 11:51, 24 September 2021 (UTC)
- The reference for the table has a note "This table copies figures unamended from a JHU (Johns Hopkins University) table". I do not think it matters whether the flags and case fatality rate are retained, but I think it is crucial that the figures are the same - as the note says - and that the default sort is the death rate. If I understand correctly you have already implemented the default sort but I am not sure about the identical figures. Dudley Miles (talk) 08:17, 24 September 2021 (UTC)
- I almost guarantee that someone will complain. But if it is difficult to include the flags, then don't bother. Since it is a complicated bot template, and not a simple table, then people will understand. --Timeshifter (talk) 00:01, 24 September 2021 (UTC)
Tol. The current scrolling template left aligns the country names via the covid classes, I believe. Inline CSS is not used for each country name.
I think though that the separate column in the sandbox template for flags messes that up. Is it possible to have the module create the regular flag plus link format together in one column, not 2? Maybe that will allow the covid classes to left align the country names.
I use sandbox 127 to see how the regular templates and the sandboxes work together. Especially to see if "show all" is working correctly for each specific table. When one table is expanded the others should go back to being partially open. Also, to be sure the covid classes, etc. are playing nice together.
I created a section to test sandbox templates:
--Timeshifter (talk) 12:00, 24 September 2021 (UTC)
- @Timeshifter: Template:COVID-19 pandemic data/styles.css uses the nth-child workaround to align columns. I'm working on a fork of those styles for this table. The whole "show all" thing is a pain to deal with; I'm not sure if it'll ever work well. Tol (talk | contribs) @ 16:31, 24 September 2021 (UTC)
- Template:COVID-19 pandemic death rates/styles.css. I prefer minimal styling; rather than add more layers to patch problems, just remove the problems. I'll work on the expand/collapse. Tol (talk | contribs) @ 16:45, 24 September 2021 (UTC)
- I've got the nth-child to work. Now I'll try expand/collapse. Tol (talk | contribs) @ 16:55, 24 September 2021 (UTC)
- @Timeshifter: I've finished expand/collapse. I used unique IDs so that it doesn't mess with other tables. How does it look? Tol (talk | contribs) @ 17:55, 24 September 2021 (UTC)
- Tol. The expand/collapse function is working correctly here:
- User:Timeshifter/Sandbox127
- But the headers are not sticky like in the other tables on that page.
- Also, there are no "V T E" links at the top left as with the other tables.
- --Timeshifter (talk) 20:21, 24 September 2021 (UTC)
- @Timeshifter: Hm; the header is sticky for me. (No headers are sticky when expanded.) The V/T/E navbar is probably unnecessary, as the template would update automatically, but I can add it if you like. Tol (talk | contribs) @ 20:30, 24 September 2021 (UTC)
- I've found the problem; I had the sticky headers gadget. I'm working on this now. Tol (talk | contribs) @ 21:22, 24 September 2021 (UTC)
Done Tol (talk | contribs) @ 21:28, 24 September 2021 (UTC)
- Tol. Thanks for the sticky headers, and the V/T/E navbar. It is useful for people wanting more info on how the updating works, and where all the data is coming from. Also, the talk page is a way to make contact, ask questions, etc.. You may never see questions asked about the template at the article talk pages for the articles it is transcluded in. --Timeshifter (talk) 21:35, 24 September 2021 (UTC)
- @Timeshifter: I see; no problem! Does it look ready to implement now? Tol (talk | contribs) @ 21:40, 24 September 2021 (UTC)
- Tol. Thanks for the sticky headers, and the V/T/E navbar. It is useful for people wanting more info on how the updating works, and where all the data is coming from. Also, the talk page is a way to make contact, ask questions, etc.. You may never see questions asked about the template at the article talk pages for the articles it is transcluded in. --Timeshifter (talk) 21:35, 24 September 2021 (UTC)
- I've found the problem; I had the sticky headers gadget. I'm working on this now. Tol (talk | contribs) @ 21:22, 24 September 2021 (UTC)
- @Timeshifter: Hm; the header is sticky for me. (No headers are sticky when expanded.) The V/T/E navbar is probably unnecessary, as the template would update automatically, but I can add it if you like. Tol (talk | contribs) @ 20:30, 24 September 2021 (UTC)
- @Timeshifter: I've finished expand/collapse. I used unique IDs so that it doesn't mess with other tables. How does it look? Tol (talk | contribs) @ 17:55, 24 September 2021 (UTC)
- I've got the nth-child to work. Now I'll try expand/collapse. Tol (talk | contribs) @ 16:55, 24 September 2021 (UTC)
- Template:COVID-19 pandemic death rates/styles.css. I prefer minimal styling; rather than add more layers to patch problems, just remove the problems. I'll work on the expand/collapse. Tol (talk | contribs) @ 16:45, 24 September 2021 (UTC)
Implementation. Automation of updates
[edit]Tol. Looks good to me. One question I have is whether the top and bottom padding in each row can be lessened. I notice that the other tables in sandbox 27 have narrower rows. So they scroll faster. Also, do we want to add {{static row numbers}}? --Timeshifter (talk) 22:06, 24 September 2021 (UTC)
- @Timeshifter: Sure. The new table uses default spacing; the others modify that. As for Template:Static row numbers, probably not. It's supposed to be small and portable, and Template:COVID-19 pandemic death rates by country (sandbox) has that for anybody who wants it. Tol (talk | contribs) @ 22:12, 24 September 2021 (UTC)
Done; I'll implement it soon unless there are complaints. Tol (talk | contribs) @ 22:19, 24 September 2021 (UTC)
- @Timeshifter, @Dudley Miles; @Gadfium: Any final comments? I plan on implementing the sandbox version tomorrow. Tol (talk | contribs) @ 23:27, 24 September 2021 (UTC)
- I'm very happy with the result. Thanks for all your work.-gadfium 23:32, 24 September 2021 (UTC)
- @Gadfium: Thank you! Tol (talk | contribs) @ 23:50, 24 September 2021 (UTC)
- I'm very happy with the result. Thanks for all your work.-gadfium 23:32, 24 September 2021 (UTC)
Done; both templates are now automatically updated. Tol (talk | contribs) @ 03:57, 25 September 2021 (UTC)
- Tol. Great work! --Timeshifter (talk) 08:26, 25 September 2021 (UTC)
- I endorse Timeshifter's comment. Great work. I have one minor comment. We must not over-estimate readers' numeracy. One reader complained that he did not understand 'Case fatality rate' - which was why I linked it - and I will be very surprised if some do not complain that they do not understand 'Deaths/m'. (They will not realise that hovering over 'm' shows million). The heading should be given in full for clarity as 'Deaths per million population'. Dudley Miles (talk) 08:48, 25 September 2021 (UTC)
- Tol and Dudley Miles. I agree. And it is more difficult in cell phones to touch the right spot. "Deaths per million" would work too. I explained "Deaths/m" in the section intro here:
- COVID-19 pandemic by country and territory#Total cases, deaths, and death rates by country (Our World in Data)
- I would rather not have to explain its meaning. And the template may be used in articles where the editors don't bother to explain.
- Can use breaks <br> or style=max-width:5em; (or whatever width works to make the header into a column with one word per line). That keeps the table narrow. --Timeshifter (talk) 09:04, 25 September 2021 (UTC)
- I think it is better to specify population as it is only an extra line and it will avoid queries. The UK government sometimes quotes proportions of the adult population. I think population is best but if it causes problems on cellphones making columns wider how about people? Dudley Miles (talk) 09:54, 25 September 2021 (UTC)
- The scrolling tables have limited vertical space in COVID-19 pandemic by country and territory. So extra header lines in the sticky headers makes for less actual table data being visible. I think the best header is this:
- Deaths
- per million
- But anything is fine by me in the end as long as it is more understandable than "Deaths/m".
- The tables will contract further in cell phones. The country names will wrap too. That is why short country names are important too. Shorter is better.
- --Timeshifter (talk) 10:35, 25 September 2021 (UTC)
- @Dudley Miles & @Timeshifter:
Done both templates (using "Deaths/million"). Tol (talk | contribs) @ 17:15, 25 September 2021 (UTC)
- @Dudley Miles & @Timeshifter:
- The scrolling tables have limited vertical space in COVID-19 pandemic by country and territory. So extra header lines in the sticky headers makes for less actual table data being visible. I think the best header is this:
- I think it is better to specify population as it is only an extra line and it will avoid queries. The UK government sometimes quotes proportions of the adult population. I think population is best but if it causes problems on cellphones making columns wider how about people? Dudley Miles (talk) 09:54, 25 September 2021 (UTC)
- I endorse Timeshifter's comment. Great work. I have one minor comment. We must not over-estimate readers' numeracy. One reader complained that he did not understand 'Case fatality rate' - which was why I linked it - and I will be very surprised if some do not complain that they do not understand 'Deaths/m'. (They will not realise that hovering over 'm' shows million). The heading should be given in full for clarity as 'Deaths per million population'. Dudley Miles (talk) 08:48, 25 September 2021 (UTC)
- Tol. Great work! --Timeshifter (talk) 08:26, 25 September 2021 (UTC)
- @Timeshifter, @Dudley Miles; @Gadfium: Any final comments? I plan on implementing the sandbox version tomorrow. Tol (talk | contribs) @ 23:27, 24 September 2021 (UTC)
Tol. Thanks. I experimented with the expanded table.
- ! style=max-width:6em; |Deaths per million people
That header would wrap to 4 lines if the max-width was lower, and if there wasn't a min-width setting for headers in the CSS:
Did you mean to set a max-width setting for the headers? That may not be a good idea either for the CSS file. I like to only use max-width for specific columns. And not to use max-width at all for the country column. I like that column to spread out in order to avoid too many laddered lines. Slower to scan tables with a lot of country name laddering. --Timeshifter (talk) 18:58, 25 September 2021 (UTC)
- (edit conflict) @Timeshifter: No; I wanted some of the smaller ones to be larger. That table is meant to be larger, and I think it looks better. I've taken the liberty to just use one-line "Deaths per million" (which will wrap if there's not enough space). What do you think? Tol (talk | contribs) @ 19:04, 25 September 2021 (UTC)
Tol. That's fine, but I think the formatting of the headers, etc. should be implemented by consensus of the regular editors of the article that the table is found in. So I think the min-width CSS for all the headers should be removed from here:
That way once you set up a bot, others can decide on the header, etc. formatting at the top. There are many options for such formatting. Check out the many tables in this article:
Sometimes narrow column heads are better. I find the tables easier to scan across. And they can compress even further in cell phones, etc. since they use max-width. It allows narrowing of the table as the browser width is narrowed. Min-width does not allow that. It can not be narrowed beyond its width. That can be a serious problem in cell phones. I never use it in tables anymore. --Timeshifter (talk) 10:26, 26 September 2021 (UTC)
- @Timeshifter: I've removed min-width. Another point — the whole point of templates is so that they can be used in many pages. In this case, it is only used on COVID-19 pandemic death rates by country, but it could be used elsewhere. Feel free to tweak the CSS; I'm good with bots, not so good with styling. Tol (talk | contribs) @ 18:14, 26 September 2021 (UTC)
Examples in separate talk section. Easier to narrow browser window in preview mode
[edit]Click the "edit source" link for this talk section only. Then click "show preview". This section is good for running tests too. Note that the top table can not be narrowed as much as the 2 tables below it without the min-width:6em on all the headers. Narrow your browser window slowly until the 2 bottom tables stop shrinking. This allows more cell phone users to see the tables.
Country | Deaths per million | Deaths | Cases |
---|---|---|---|
test | test | test | test |
Country | Deaths per million | Deaths | Cases |
---|---|---|---|
test | test | test | test |
Even narrower:
Country | Deaths per million | Deaths | Cases |
---|---|---|---|
test | test | test | test |
--Timeshifter (talk) 13:03, 26 September 2021 (UTC)
Documentation changes
[edit]Hello @Dudley Miles, you reverted my documentation changes so I'm bringing them up here. Hatnotes should go above mboxes (in articles per Wikipedia:Manual of Style/Layout#Order of article elements and elsewhere by convention). As for the hatnotes, they should be formatted using the standard Template:Hatnote. I removed the sentence on being bot-updated because there's already a link to the Lua module and it's in a category; it should probably be added in the body text. I also removed the link to Template:COVID-19 data/Cite because it just contains the reference found in the table, and formatted the other links using Template:Hatnote. Could you please explain your reversion? Your edit summary of "correct before" is not helpful — it does not matter whether something was good, it matters whether it's better now. Tol (talk | contribs) @ 14:07, 13 October 2021 (UTC)
- I was over-hasty reverting, but the note of updating by a bot should be there for a reader who looks at the table even if it is also elsewhere. How about going back to your version with the addition of the bot updating - and the deletion of the blank reflist? Dudley Miles (talk) 14:25, 13 October 2021 (UTC)
- Sure, with the exception of the reflist — it's there for the main template page's reference (Template:COVID-19 pandemic death rates#References). Tol (talk | contribs) @ 14:34, 13 October 2021 (UTC)
Has updating stopped working?
[edit]Hi Tol. Has auto-updating stopped working? The figures have not been updated for the past two days and the ref is no longer showing the date of update. Dudley Miles (talk) 12:56, 13 November 2021 (UTC)
- @Dudley Miles: Nope, the citation template was broken with a recent edit. I've reverted it for now. Tol (talk | contribs) @ 14:27, 13 November 2021 (UTC)
- Thanks. Dudley Miles (talk) 14:30, 13 November 2021 (UTC)
- No problem! Tol (talk | contribs) @ 14:41, 13 November 2021 (UTC)
- Thanks. Dudley Miles (talk) 14:30, 13 November 2021 (UTC)
Slovakia cases
[edit]Slovakia in fact has 593,242 total confirmed cases of COVID-19[1]. The template currently shows 1,040,996 cases. Please correct it. --2A02:AB04:2AB:700:7C98:CD33:4BD2:9881 (talk) 17:37, 19 November 2021 (UTC)
References
- The table takes its data from Our World in Data. We do not correct its figures. Dudley Miles (talk) 19:52, 19 November 2021 (UTC)
However, Our World in Data is definitely wrong in this case. (Edit: I have already contacted the Our World in Data page about the problem. I wonder where they got this data from.) --2A02:AB04:2AB:700:49A1:51D4:A125:3E49 (talk) 21:50, 20 November 2021 (UTC)
- I am glad you were able to contact them. We used to use Johns Hopkins University data and I never managed to find an address to contact them about errors. Dudley Miles (talk) 22:51, 20 November 2021 (UTC)
Country wraps on Firefox wide screen
[edit]@Tol: Per the "21 December" part of this talk, in Windows Firefox the longest country "Saint Helena, Ascension and Tristan da Cunha" wraps on a wide screen (no horizontal scroll needed), which doesn't occur in Chrome. You can fix this by adding overflow-block: scroll;
. Also, "height" doesn't appear to be needed and may hinder the responsiveness of "max-height". Finally, you can get rid of "display" if you change "max-width" from "100%" to "max-content", which maintains a vertical scrollbar close to the table.
Template:COVID-19_pandemic_death_rates/styles.css
#wrapper65150380 {display: inline-block;max-width:100%;max-content;height: 50em;max-height: 75vh; overflow: auto; overflow-block: scroll; }
Jroberson108 (talk) 08:36, 23 December 2021 (UTC)
- @Jroberson108: I'm by no means good at CSS — you can make the changes yourself if you think they would be helpful. I don't really know how
max-width: max-content;
oroverflow-block
work. Tol (talk | contribs) @ 04:20, 24 December 2021 (UTC)- @Tol: I applied the changes. For some reason the CSS sanitation script identifies "overflow-block" as
Unrecognized or unsupported property ...
, so I applied it inline in the template. I tested on Firefox and Chrome and the text doesn't wrap anymore. I see there is an ongoing discussion here about including "unsupported" styles: T162379. Any additional info that might help? Jroberson108 (talk) 05:50, 24 December 2021 (UTC)- @Jroberson108: It looks good to me! I'm not familiar with TemplateStyles/css-sanitizer, so I'm guessing that's just because it's a new enough property that it hasn't been whitelisted yet. Thanks for your changes! Tol (talk | contribs) @ 18:59, 24 December 2021 (UTC)
- @Tol: I applied the changes. For some reason the CSS sanitation script identifies "overflow-block" as
World deaths per million incorrect?
[edit]@Tol: The World's "deaths per million" seems to low at 722. Jroberson108 (talk) 07:24, 3 February 2022 (UTC)
- That is what's in the JSON data, so maybe an issue with the source? Jroberson108 (talk) 07:26, 3 February 2022 (UTC)
- @Jroberson108: I don't think so; the the data the same as OWID's, whose calculations appear correct. For and , . Tol (talk | contribs) @ 00:12, 4 February 2022 (UTC)
- @Tol: Shouldn't "World" be a summation of its parts? There are many countries under it that report higher deaths. Jroberson108 (talk) 03:42, 4 February 2022 (UTC)
- @Tol: For example, World currently shows: 723 | 5,700,923 | 384,859,233.
- When I do a sum excluding World, I get: 218,044 | 6,661,687 | 475,079,086.
- Jroberson108 (talk) 04:21, 4 February 2022 (UTC)
- The 723 is right according to the equation, but the other two seem off. Jroberson108 (talk) 04:25, 4 February 2022 (UTC)
- Jroberson108. You are including the European Union in your sum. In the Visual Editor when I select the whole column of death numbers except for World and EU, a sum pops up immediately: 5.700,908. See where I copied the expanded table in order to do this in VE:
- User:Timeshifter/Sandbox172
- --Timeshifter (talk) 12:06, 4 February 2022 (UTC)
- @Timeshifter: Ok, thanks. In that case, should the EU also use "sorttop"? Jroberson108 (talk) 12:18, 4 February 2022 (UTC)
- Jroberson108. I like it to sort. Because I like to directly compare the US to the EU in the rankings. --Timeshifter (talk) 14:51, 4 February 2022 (UTC)
- @Timeshifter: Ok, thanks. In that case, should the EU also use "sorttop"? Jroberson108 (talk) 12:18, 4 February 2022 (UTC)
Updating has stopped
[edit]Tol the table has not updated since 12 September. UK deaths are shown as 205,718 but OWID shows them as 206,150. Can you take a look please. Thanks. Dudley Miles (talk) 11:33, 16 September 2022 (UTC)
- @Dudley Miles, the table currently lists UK deaths as "206,150" (the bot updated the data shortly after your post). It looks like the bot has been updating the data daily as it's supposed to (see data history), but the UK data wasn't updated in the OWID file for some reason. Tol (talk | contribs) @ 19:34, 16 September 2022 (UTC)
- Thanks Tol. However, it was not an issue with UK data. The table was showing 12 September at the top of the page for four days as the last update, so either the update was unsuccessful on the intervening days or the update date on the page was not being changed. Presumably it was a temporary issue which has now resolved itself.
- UK data is updated erratically, but that is a separate issue probably because the UK no longer provides daily updates. Dudley Miles (talk) 08:31, 17 September 2022 (UTC)
- @Dudley Miles: Hmm. I don't know why that would be. It looks like the data and the module are both working fine. Perhaps try purging the page if this happens again? Tol (talk | contribs) @ 18:36, 17 September 2022 (UTC)
- UK data is updated erratically, but that is a separate issue probably because the UK no longer provides daily updates. Dudley Miles (talk) 08:31, 17 September 2022 (UTC)
Font for "European Union" in Italics ?
[edit]To whom this may concern (everyone reading this template), the font for the "European Union" should be in italics, if at all possible, since it's not simply another country like the others in the top numbers of Covid-19 infections. It simply would "look better" that way to the reader. How does one do that, please? Thank you. User : Le Anh-Huy (https://en.wikipedia.org/wiki/User:Le_Anh-Huy)
Error message on template output
[edit]"Lua error in Module:COVID-19_data at line 149: attempt to concatenate field 'name' (a nil value)." --Timeshifter (talk) 06:17, 10 March 2023 (UTC)
- @Timeshifter: I fixed the error in the module. Virgin Islands (VI) and Puerto Rico (PR) are missing the "name" property in the JSON data, which can be seen in the table. @Tol: may have to take a closer look as to why the data is missing. Jroberson108 (talk) 10:58, 10 March 2023 (UTC)
- Thanks Jroberson108! --Timeshifter (talk) 11:08, 10 March 2023 (UTC)
- @Jroberson108, @Timeshifter: Sorry; I haven't been too active recently. Thanks, Jroberson108, for the patch. It seems like data was recently added for many smaller territories/entities, including some which hadn't been fully set up (Virgin Islands, Puerto Rico); this broke some stuff which tried to find a name which wasn't there. I'm adding names to them now and reverting the patch, which should be no longer necessary. Tol (talk | contribs) @ 06:05, 14 March 2023 (UTC)
- Thanks Jroberson108! --Timeshifter (talk) 11:08, 10 March 2023 (UTC)