Template talk:COVID-19 pandemic data/United Kingdom medical cases chart

Official numbers
The chart is supposed to be based upon the official 1400 numbers, so why is 11 March 460 here, when the official number was 456? Dan100 (Talk) 14:30, 12 March 2020 (UTC)

Ah someone has just corrected it. Good show Dan100 (Talk) 14:32, 12 March 2020 (UTC)


 * To clarify, as Dan100 has said the numbers in this chart are based on official PHE figures released once a day corresponding to confirmed cases at 0900 that day. There are reporters giving different figures during each day, which makes this look behind, but it is not a current count of the latest figure reported. Please do not edit previous days figures unless PHE adjust them retrospectively, and please do not update this based on any other information. Thanks |→ Spaully ~talk~ 18:55, 18 March 2020 (UTC)


 * Wouldn't it be more straight forward if the numbers matched those in the PHE graph? It shows 3605 as the total toll as of today, whereas this page shows it as yesterday's number. Bendž|Ť 19:57, 3 April 2020 (UTC)


 * Although it seems strange that PHE are recording the deaths a day late, if I understand correctly? Bendž|Ť 20:09, 3 April 2020 (UTC)
 * Yes they are reporting deaths up to 5pm the previous day which is why this changed on this chart. I think this more accurately represents the data if comparing cases to deaths, and comparing the death numbers before/after their change in reporting. Though it's interesting to see their choice of graph, it's worth a discussion. 12:55, 4 April 2020 (UTC)
 * You're right, it's more accurate this way. Bendž|Ť 13:18, 5 April 2020 (UTC)

Chart needs widening
Anyone know how to do it? Dan100 (Talk) 14:42, 12 March 2020 (UTC)
 * You can control the width of the bar area with  and the width of each number with  . Check the docs: Medical cases chart. Alexiscoutinho (talk) 17:36, 12 March 2020 (UTC)

WikiProject COVID-19
I've created WikiProject COVID-19 as a temporary or permanent WikiProject and invite editors to use this space for discussing ways to improve coverage of the ongoing 2019–20 coronavirus pandemic. Please bring your ideas to the project/talk page. Stay safe, -- Another Believer ( Talk ) 18:18, 15 March 2020 (UTC)

Recoveries
This information is becoming increasingly stale - it clearly can't be true that there are only 18 recoveries and we are otherwise using only PHE data which they are not producing for recoveries. I think we should remove recovery information from this graph until this information can be more reliably sourced. Any thoughts? |→ Spaully ~talk~ 20:30, 16 March 2020 (UTC)
 * Some wrong recovery figures have been added. The 'Total Recovered' panel here  in fact shows the total number as of last count across the UK, Crown Dependencies and Overseas Territories.   and others, please note.  The UK government hasn't published a recovery figure since 22 March.  See also this discussion.  As such, I've removed the recovery figures from the graph from 23 March onwards.  I'll add a note. — Smjg (talk) 12:02, 8 April 2020 (UTC)

March 18 Number of Deaths
Hey all. I just noticed that the chart's value is 104 deaths, while the number presented on the right-hand side is 103. I'm not sure which is which but this should probably be sorted out. Bluegreenmagenta (talk) 23:12, 18 March 2020 (UTC)
 * Yes check.svg Done Thanks |→ Spaully ~talk~ 09:36, 19 March 2020 (UTC)

March 18 and 19 deaths are still wrong; at least the increase is anyway. 81.156.182.106 (talk) 22:51, 25 March 2020 (UTC)

Change of recording/reporting
Just a few things:


 * The Department of Health and social care have changed the way they're recording and reporting deaths. I feel like that should be noted in the notes section but not quite sure how to write that myself so if someone could do that, that would be great.


 * I think that the chart may actually be one day ahead as the reports we have been following have actually been for the previous day, but will obviously need more people to check the departments explanation from their official twitter account to clarify that.


 * The DHSC have also said that the figures we currently have down as March 25 although they are actually for March 24, did not actually cover the whole 24 hours and only covered 9am to 5pm.

If someone could check the tweets from DHSC as the chart will need a few changes from my understanding. --Brandon Downes (talk) 19:13, 26 March 2020 (UTC)
 * Thanks for spotting that. I have corrected the numbers and added a note about the change. Feel free to edit/adjust. I have backdated the death numbers for 24th and 25th, which leaves 26th empty; this is the correct way to do it but I bet will cause confusion and people to edit it thinking they are correcting it. BW |→ Spaully ~talk~  09:27, 27 March 2020 (UTC)
 * Yeah, will probably be best to get some edit permissions set up as this is going to be a daily occurrence of people editing when they don't understand the change in reporting. Also from what I understand the reported cases should still be moved 1 day back as you can't have the data for a day before the day has ended. For example we currently have March 27 data for cases as 14,579 but shouldn't this be the data for March 26? Just need some clarification as I don't know if I'm not understanding, but to me the way we currently have the graph doesn't really make sense. --Brandon Downes (talk) 15:01, 27 March 2020 (UTC)

Excessive data on the chart
The change in deaths should either be absolute or relative as having both bloats the chart (this is not a spreadsheet) and breaks the lines on mobile. I'll wait until a decision is made to convert the chart to the newest standard ('data' parameter). Alexiscoutinho (talk) 03:24, 27 March 2020 (UT
 * I agree, there is no particular interest in having both. Percentage change is more informative as the absolute increase can be worked out at a glance. appears also to agree given his edit comment. |→ Spaully ~talk~  09:05, 27 March 2020 (UTC)

Please could someone fix my editing.
Dear all,

I have tried adding a footnote that explains some extra tests that were added to the figures today, but I can’t get the footnote to work. I just want the note to appear underneath the existing one. Also, could someone source this footnote please, I think you’ll find the info on the U.K. Government website. Thank you. — Preceding unsigned comment added by Pablothepenguin (talk • contribs)
 * I think someone has fixed this now, thanks. |→ Spaully ~talk~ 20:25, 14 April 2020 (UTC)

Added value of showing the number of patients who have recovered
Within the figures for Confirmed Cases, it would be most helpful to show the number of patients who have recovered (discharged from hospital). Over time, I hope this would become a reassuring number (proportion) for many readers of Wikipedia. More importantly, this would give a visual indication of the number currently hospitalised (effectively subtracting from the total those who have regrettably died and those who have recovered). This is a key number to compare with the number of ICU beds available in the UK. — Preceding unsigned comment added by 185.13.50.186 (talk • contribs)
 * Unfortunately the UK government is not currently collecting/releasing this information currently and so there are no up to date numbers. This is an important number I agree to add when available. |→ Spaully ~talk~ 20:25, 14 April 2020 (UTC)

Date of deaths
Hi,

This chart seems to be misleading - the date that deaths are announced is the date when they are processed, not the date they occurred. See for instance the dates that the deaths occurred which were announced today (England only) at https://ichef.bbci.co.uk/live-experience/cps/624/cpsprodpb/vivo/live/images/2020/4/25/b2bc5b1a-c99b-42da-9204-a70ac7868b62.png

Personally I would have said this was a far more accurate representation, showing deaths by date announced isn't the most accurate metric.

Thanks in advance,

Best wishes, 2A00:23C7:E1C:B600:A84E:6598:5BE1:10A2 (talk) 14:30, 25 April 2020 (UTC)
 * If you can find a link to a reliable source that provides this data and is kept up to date, then we could create something from it, though likely to be a separate graph uploaded as an image. I agree that is a very useful graph and is more useful for assessing progress, though takes more work and technical capability than keeping this up to date with PHE figures. |→ Spaully ~talk~ 18:17, 25 April 2020 (UTC)
 * Like you suggest that's a bit trickier to find for the UK as a whole. I think it will need data from the four home nations to be added together as it is available for each of them individually: this NHS England data; Scotland logs them in this page I think, which also links to this weekly roundup with deaths by date on p12; I believe this info showing Wales deaths is by date of death (although there may be methodological differences to other home nations as it includes some care home deaths); and for NI this weekly report (p2) found here contains the info.


 * Best wishes, 2A00:23C7:E1C:B600:FDC8:FC76:9922:E03D (talk) 06:41, 27 April 2020 (UTC)

Places not included
People keep reinstating the specific reference to care homes in the note about what the death counts do and don't include. Why?

The death counts don't include deaths anywhere other than hospitals. As far as I can see, there's no reason to single out care homes for mention here of all the many settings other than hospitals in existence. It may be true that the media are doing it frequently, but that doesn't mean we should copy them. — Smjg (talk) 22:13, 25 April 2020 (UTC)
 * It's gone through quite a few iterations and I think we need to keep it concise and of course correct. It is in the media, and there are reports that there may be a significant proportion of total deaths occurring in care homes so that clarification seems pertinent. This is an explicit point made on the PHE page: These figures do not include deaths outside hospital, such as those in care homes, except as indicated above.. |→ Spaully ~talk~ 14:48, 26 April 2020 (UTC)
 * Maybe, but the statement doesn't say that, and thus it reads as just an arbitrary example of an environment that isn't a hospital. It's true that the majority of deaths outside of hospitals seem to be in care homes, and that according to some sources these account for anything up to a third of all coronavirus deaths.  Thinking about it now, I suppose what's needed is to amend it so as to capture this (with sources of course).  Maybe I just need to go away and think about it a bit more. — Smjg (talk) 20:32, 27 April 2020 (UTC)


 * Now that the official death count has been amended to include all settings, I've made a few updates to reflect this. At the end of the lead in the main article: "It has been feared that there have been several thousand further deaths in the community, and especially in care homes, which until 29 April were not included in the official government figures."  I think this suffices to indicate that the majority of deaths outside hospitals are in care homes, and there's no need to repeat any explicit reference to care homes in this box.
 * As an aside, there's an anomaly in that the first all-inclusive count (hereafter referred to as AIC) is shown alongside the date 28 April, but everywhere else it's said that these AICs started on 29 April. I make out that this first AIC covers deaths up to 17:00 on 28 April, and 29 April is merely when this figure was published.  So the wording needs to be amended a bit so that the article doesn't seem to contradict itself.  If anyone wants to take a stab at it then please feel free; otherwise, I might do so later when I've a bit more time. — Smjg (talk) 11:06, 30 April 2020 (UTC)

Button processing code
Can someone knowledgeable on the processing required for the Jan, Feb, Mar and so on buttons please check this. I added a button for May, and it seems to work, but needs reviewing. Thanks. Arcturus (talk) 17:27, 1 May 2020 (UTC)

Muddled-up numbers
As I'm guessing quite a few people have realised, the latest data on gov.uk contradicts the numbers we have here. As of right now: which of course doesn't match up with 248,818 being the figure we have for 19 May. Does anyone have any initial thoughts about how we can reconcile the discrepancy?

(It's furthermore annoying that the historical tallies by the various measures are all over the place and, in some cases, nowhere to be seen at all. It would be nice if there were a CSV download with all the figures as of each date in.) — Smjg (talk) 08:18, 21 May 2020 (UTC)


 * yes I noticed that. I have finally worked out how to edit the chart, but although the figure for deaths tallies, that for cases doesn't. Unfortunately it seems impossible to access the gov.uk update for previous dates. I did however amend the case increase percentage for 19 May to 1 decimal place (ie 1.0% not 0.98%) so it is now consistent with everything all the other figures. Saxmund (talk) 08:26, 21 May 2020 (UTC)


 * There is a further problem in that, unless we can find the published historical figures, we can't meaningfully source the figures other than the latest one.  On 13 April you updated the chart with backdated counts of confirmed cases including tests on key workers.  Where did you get these numbers from? — Smjg (talk) 17:42, 21 May 2020 (UTC)
 * And can someone please explain how the reallocation of test results back in time can reduce the cumulative number of cases? — Smjg (talk) 22:13, 21 May 2020 (UTC)
 * It was from gov.uk, which had a list of key worker tests at the time (see the archived version of the web page on 13 April). If you want more archived versions on that web page, see the Wayback Archive. -- Ted Edwards  01:35, 22 May 2020 (UTC)
 * Here's the garbled explanation from GOV.UK; Daily totals reflect actual counts reported for the previous day. Each day there may be corrections to previous reported figures. This means that previously published daily counts will not necessarily sum to the latest cumulative figure. It also means that today’s cumulative count may not match the previous day’s cumulative count plus today’s daily count. It is a problem that each day's update appears to be overwritten by the following day. Archived versions of the page would need to be accessed as sources for discrepancies, but since the figures appear to be continuously updated and backdated, this could be problematic. Arcturus (talk) 10:46, 22 May 2020 (UTC)

Continuing irregularities in case data
The source data is issued by UK Government on a daily basis (usually a few hours late from their stated time). It comes form this website. Here is an extract from the notes pertaining to 26 May 2020:

Due to technical difficulties with pillar 2 data collection we cannot provide people tested figures today.

''There may be a small percentage of cases (around 2%) where the same person has had more than 1 positive test result for coronavirus. The government will correct any figures if they have subsequently been found to have an error''

''The number of daily tests completed is 38,682 lower, lab confirmed cases is 2,039 lower, and in-person tests is 36,227 lower than the difference between today’s and yesterday’s cumulative totals. This is due to the integration of the Cambridge laboratory into Pillar 2’s Management Information, which began operating on 11 May, reporting of Milton Keynes lab testing activity from the 24 May, which was not reported yesterday under Pillar 2 due to IT issues, and other revisions to historical data across all Pillars due to further information being made available to support data processing.''

This text will be overwritten tomorrow (27th) and I think will only then be available via some web archive tool, if at all. The overall protocol is - not to put too finer point on it - crap. Of particular concern is the last part of the statement - "other revisions to historical data across all Pillars due to further information being made available to support data processing". I assume this means they update some of the daily historical data; but that updated data is not readily available, as noted above. As a consequence, some, maybe many, of the daily case figures will be inaccurate. The notes given above in italics are typical of most days, where one or more anomalies are described. I'll have a look for a better source, but I'm not hopeful. Maybe Johns Hopkins data might be better. Maybe we should put a disclaimer in the template. Any ideas as to how best to proceed? Arcturus (talk) 20:16, 26 May 2020 (UTC)

Hi, could someone please update this. — Preceding unsigned comment added by 2A02:C7F:3A91:9800:4D2E:4E25:5B01:CA7F (talk) 20:37, 28 May 2020 (UTC)

Number of deaths
The column for number of deaths seems to have been shifted by a day. All the figures given should have been a day later. See for example the official source - - in the graph for "Total number of COVID-19 associated UK deaths by date reported" at the bottom, the total number of deaths in a particular day can be seen when you hover over the points in the curve. The first death recorded was for 6 March (not 5 March given in the chart), the latest one for 29 March was 38,161, not 28 March. Hzh (talk) 11:40, 30 May 2020 (UTC)

Also found one minor discrepancy - the number for 9 May is 31586, not 31587 given for 8 May. Hzh (talk) 11:51, 30 May 2020 (UTC)


 * What happened with the figure for 27 June? On 26th, 43,514 dead, on 27th 43,350 and that is not an increase of 0.08%.Davidbstanley (talk) 17:31, 28 June 2020 (UTC)
 * The death figure issued today, but for yesterday, needed correcting - just a typo. The .GOV website gives the death figure for a day in arrears, at least according to this statement; As of 5pm on 27 June, of those tested positive for coronavirus in the UK, across all settings, 43,550 have died.. The .GOV website data can be difficult to interpret. See the comments I made above, and have a look a the explanatory notes on the website; it's not easy to follow what they are doing. Arcturus (talk) 18:55, 28 June 2020 (UTC)

I have moved the death data back to the "previous" day. The two sources quoted continue to state that the death figures are accurate to 5pm the previous day, whereas tests and new cases are up to 9am that day. I don't see any reason for a change of policy between 1 and 2 July, particularly as no note has been given to explain it. The policy of quoting deaths a day behind is consistent and inline with the data. It is true however that https://coronavirus-staging.data.gov.uk/deaths is now quoting the number of deaths against the day reported. However the pages we quote as the source haven't changed their practice Saxmund (talk) 19:28, 5 July 2020 (UTC)

We stopped reporting the death data on 16 July, when Public Health England stopped issuing the data after it came to light they'd been deliberately exaggerating the figures. However, the daily counts are still available here. This is a beta site from .GOV.UK and the death figures can be found under "More detail" in the Death section. I assume the death figures are still being inflated as before; it would seem so based on this comment (my emphasis); ''Data on deaths from these 3 sources are linked to the list of people who have had a diagnosis of COVID-19 confirmed by a PHE or NHS laboratory (pillar 1 of the government's mass testing strategy) or through testing in commerical [sic] labs (pillar 2). This is to identify as many people with a confirmed diagnosis who have died as possible.''. The full explanation is available here. Do you think we should include these data? If so, I will backdate it in the template. Arcturus (talk) 15:40, 28 July 2020 (UTC)
 * My instinct would be not to report data if it is questionable. And the death figures are certainly questionable by the governments own admission.  Of course the cases figures are questionable too, but that is the same for every country. Davidbstanley (talk) 06:07, 29 July 2020 (UTC)
 * I have added the latest death figures, as on https://coronavirus-staging.data.gov.uk/ the death figures are still being updated. I know the death figures are questionable, but so are all of the death figures from before 17 July and they are still on the template. Because of that, I didn't see any reason not to add the latest figures. I definitely think adding a note explaining how the government calculates the death figures would be very helpful. But that doesn't explain why Public Health England said they would stop issuing death data on 16 July, yet the official site continues to be updated. DarkHorse234 (talk) 17:52, 29 July 2020 (UTC)

Some editors are adding the death figures, others aren't. I didn't add today's, but if anyone feels that we should, please do so (65 deaths - ). My view at the moment is that we shouldn't. The England element of the UK death figures are still being falsified by including deaths of anyone who ever had CV19 even if they recovered from it. Arcturus (talk) 16:11, 5 August 2020 (UTC)

July 24
July 24 is missing, can someone please add it? — Preceding unsigned comment added by 178.11.117.167 (talk) 07:20, 27 July 2020 (UTC)
 * Done, and daily data brought up-to-date. Arcturus (talk) 15:27, 28 July 2020 (UTC)

Revised death toll at 41,329
I understand the reason for this, but it is nor clear toy me how we post date these?
 * For now, I have just added a note explaining the change. I haven't found any death data with the revised conditions yet. DarkHorse234 (talk) 14:33, 13 August 2020 (UTC)

Percentage or absolute option for new infections?
I started a discussion on that topic here Project COVID-19, Medical cases charts - change type. My personal opinion is that absolute change display suits much better in an epidemic's second wave. But also the status quo ante (percentage) should be kept here until some agreement is reached (locally or possibly a consensus in linked discussion). -- Kohraa Mondel (talk) 15:28, 3 December 2020 (UTC)
 * The above-mentioned discussion has now got lost in the archives (I never got around to contributing to it). However, let's perhaps restart it here, with particular reference to the UK chart. Compare what we have at the moment with the German chart: Template:COVID-19 pandemic data/Germany medical cases chart. To me, the German chart is superior. It conveys the day-to-day situation much more readily than the current UK chart. Should we move to absolute values? Arcturus (talk) 21:37, 4 January 2021 (UTC)

In addition: some minutes ago the screen in the corner (BBC) pronounced today's "record" figures, no per cents...~, are they ?&#32;— Pietadè (talk) 16:45, 5 January 2021 (UTC)
 * Support move to absolute values. Arcturus (talk) 21:37, 4 January 2021 (UTC)
 * Support move to absolute values. If I open this table side by side with a table where absolute values are displayed, the latter seems to me (far) more informative for UK (population ~67,886,004).&#32;— Pietadè (talk) 08:04, 5 January 2021 (UTC)
 * Agreed, the percentage now is not useful information and absolute daily rise in cases/deaths seems sensible. |→ Spaully ~talk~ 13:42, 5 January 2021 (UTC)
 * Support. As often stated before, in my opinion percentage makes only sense for cases where active cases outnumber inactive cases by far. So here absolute differences should be used for cases as well as for deaths. Only acitive cases can infect, and only active cases can die. Comparing to the total number of cases, where most of them are inactive, never makes sense. -- Kohraa Mondel (talk) 19:12, 30 March 2021 (UTC)

are you able to help, or advise, on how this modification can be implemented? Thanks, Arcturus (talk) 21:36, 21 January 2021 (UTC)
 * If you want to keep the old percentages, then the only way I see is to start writing changes manually. I don't think implementing code for multiple change types in one column is the right approach. Especially because splitting daily and cumulative counts (and their attached changes) is planned in the near future. Alexiscoutinho (talk) 23:50, 21 January 2021 (UTC)
 * I think the suggestion was to have the display as shown here, where the Death figures are given in absolute terms, but cases remain as percent (, is this your understanding). The old percentages would not need to be kept in the Death data. Is this feasible, and relatively easy to do? Thanks, Arcturus (talk) 19:32, 31 January 2021 (UTC)
 * Today's: "deaths = +0.5560239080808176% | cases = +0.5555192608812019%", where the first refers to "+21,088" cases and the latter to "+587" deaths, seems to support the way of presenting as «cases (+0.???%)» «deaths (+ number of dead)», at first glance &#32;— Pietadè (talk) 19:51, 31 January 2021 (UTC)
 * By the way, just now (19:04, 1 February 2021 (UTC)) changed „|changetype1= a“ to „|changetype1= p“ in etwi version, any thoughts? &#32;— Pietadè (talk) 19:04, 1 February 2021 (UTC)

Is this by any means what you intended to do? Scrubbed it down (mostly regexp) to redundance free data version using the lua module with automatic calculation. It's somewhat restricted where it comes to column header texts, though. -- Kohraa Mondel (talk) 18:03, 28 March 2021 (UTC) P.S.: Sorry it messes up the discussion a bit. I think I should remove it as soon as you have decided whether to take over this example or parts of it, or not. -- Kohraa Mondel (talk) 18:05, 28 March 2021 (UTC)


 * I think so. You can have any column header you want if you specify  and  . By the way, wow! I was so lost in my first reply. What you wanted was much simpler than I thought. I even feel like I killed the discussion... Alexiscoutinho (talk) 23:24, 28 March 2021 (UTC)
 * Ok, fixed it, thx! Let's see if the others actually intended the single change type per column... -- Kohraa Mondel (talk) 00:38, 29 March 2021 (UTC)

Now that percentage in cases change is very low and thus has a high relative rounding error, I'd change to absolute daily change mode there, too. Agreed? -- Kohraa Mondel (talk) 16:01, 4 May 2021 (UTC)
 * Yes, 'cause (e.g., 0.000001...% doesn't appear any way to be informative).&#32;— Pietadè (talk) 16:46, 4 May 2021 (UTC)
 * The template actually rounds to two fraction digits, so currently there is only one meaningful digit apart from leading zeros. 0.000001% would be rounded to 0,00% and then replaced by '='. So I'll change now to absolute...-- Kohraa Mondel (talk) 15:55, 5 May 2021 (UTC)

Backup
Copying hereby links to data published at coronavirus.data.gov.uk, as of 18:48, 4 December 2020 (UTC), at least FF opens the files:

Deaths / Cases, in case something changes (vanishes); by the way, dates/cases/deaths published there are not in correspondence with the ones published in this table.&#32;— Pietadè (talk) 18:48, 4 December 2020 (UTC)

Why are confirmed cases red, as well as deaths?
This is a little misleading; at a glance, I just see red, and have to consciously look to see that deaths and confirmed cases are different shades of red. They should be different colours - red for confirmed cases and black for deaths, for instance. — Preceding unsigned comment added by 92.7.22.149 (talk) 21:26, 26 January 2021 (UTC)

Best Attempt at Adding Recoveries and Active Using "Currently Hospitalized" from OurWorldInData
Below is a graph of my attempt at adding in the recoveries and active cases info per the title Araesmojo (talk) 18:04, 9 August 2021 (UTC)

JSON format
In JSON format (updated daily): cases / deaths &#32;— Pietadè (talk) 19:37, 9 August 2021 (UTC)

Using 7-days incidence for changetype in cases and deaths
Seven days incidence is now available from the medical cases chart module. To me it's much better to depict the current epidemic situation and makes for a much better comparability between countries. It could look like this:

-- Kohraa Mondel (talk) 19:24, 7 November 2021 (UTC)