COVID-19 Canadian Dashboard

Canadian Covid 19 Dashboard Social isolation can really change one's behavior. For me, it has meant spending a lot more time compiling and exploring different datasets in R. As I'm sure many others are, I have recently been playing with COVID-19 data. I was curious about how my home country of Canada has been handling COVID-19 so I created this dashboard to monitor the progress. Each graph hosted here is generated by my publicly available R code1 and updates whenever I run the code (which tends to be daily). The data comes from official Canadian government reports2. If you have any questions, do not hesitate to ask.

A few notes on the data...
As one might expect, data during a pandemic isn't always accurate and up-to-date. There are several considerations when interpreting this data:
  • New cases and deaths are reported on a 14-day rolling average.
    • As day-to-day statistics can be volitile and influenced by systematic varaibles, I used a rolling 14-day average to aide in understanding the long-term trends of COVID-19. The most notable effect this has on the graphs is it makes the line of fit smoother rather than looking like a heart-rate monitor.
  • The Canadian Territories are not included in any provincial breakout graphs.
    The territories have had very few cases of COVID-19. However, the territories are included in any aggregate Canada statistic (cases, deaths, etc.).
  • Some Graphs Exclude Québec.
    Particularly early on, Québec had an abnormally high COVID-19 incident rate. Graphing these high rates often obscures what is happening in other provinces. As such, I provide an option to see graphs excluding Québec. However, Québec is always included in any aggregate Canada statistic (cases, deaths, etc.).
  • On occasion, this government data has errors or time lags in some metrics.
    These issues can affect the graphs. For instance, several times the daily data has been uploaded with up-to-date "total cases/deaths," but no data on "new cases" have been uploaded (sometimes their math does not quite add up). These issues can cause some graphs to report up to different time points or the most recent data to be slightly off.
  • Generally speaking, the most detailed and up-to-date information is local data.
    However, while these graphs do not report same-day data, the most recent few days might not be completely accurate either due to delayed reporting.
  • Provincial confirmed cases might differ due to testing or reporting practices rather than how prevalent the virus is.
    These data come from thousands of medical facilities, hundreds of municipalities, and 10 provincial governments. The pandemic response, reporting, and data management are not consistent across these diverse administrations. As such,
  • These data only represent cases the government is aware of.
    This includes individuals who tested positive for COVID-19 as well as probable cases. However, there are likely many more cases not reported or known about.
  • This data does not include any information on COVID-19 recovery.
    As such, any reference to total cases does not refer to ongoing cases, but rather the raw number of cases ever reported since the beginning of the outbreak.
  • The "death rate" calculations might be misleading.
    The calculation is simplistic--total deaths divided by total resolved cases (recovered + deaths). As such, this rate is likely inflated by the fact that cases are certainly underreported and as such might not be a fair between provinces or countries depending on testing prevelance and COVID death critera. As a note, ongoing cases are not taken into account and some data suggest the time from symptom onset to death ranges from 2 to 8 weeks3.

DOWNLOAD THE DATA
DOWNLOAD THE R CODE

Summary Table




Summary Graphs



*Please note the differing scales. The scale for deaths is 10x the scale for cases.

COVID-19 Cases













COVID-19 Deaths













COVID-19 Death Rate




*COVID deaths as a percentage of resolved cases. This data might not be what you think it is as total COVID cases are likely underreported. This, along with differences in what qualifies as a COVID death make comparisons between provinces or countries difficult and perhaps misleading.

Active COVID-19 Cases







1Government of Canada (2020, April 2). Coronavirus disease (COVID-19): Outbreak update. https://www.canada.ca/en/public-health/services/diseases/2019-novel-coronavirus-infection.html 
2Jensen, L. (2020, April 2). Canadian COVID-19 Cases.  https://drive.google.com/uc?id=1hcUw6TIUKKMfZMKn93PqlpPnX1gzZg1d
3The World Health Organization. (2020, February). Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19)  https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf

Comments

Post a Comment

Popular posts from this blog

What percent of my social media content are ads?

Image
I collected this data by using each companies’ official android app. Using the home page of the app, I scrolled at a consistent moderate speed for the first 100 posts/elements and counted how many elements were advertisements or suggested content. Advertising behavior is a fickle thing and might change depending on how frequently you use the app, how and what advertisements you engage with, or even how quickly you scroll through the information. As such, your mileage may vary, but these were my personal results—take them for what they are worth.  I made the graphic in Excel and PowerPoint. I have provided more information about the collection for each app. Facebook: I use Facebook lite under 30m a day but nearly every day. A few times when scrolling (2-3/100 posts), Facebook “suggests” content or posts where I have not liked anything from the page nor had my friends engaged with it. I counted these suggestions as advertisements. Instagram: I use Instagram a few times a month and ...
Read more

Boy vs Girl Marvel Movies?

Image
Purpose: To investigate what, if any, Marvel (MCU) movies males prefer more than females.     Source(s): I used the Wikipedia list of MCU films  Wikipedia page . I scraped most IMDB ratings mid-2019, but manually added in the post-2019 movies last week. When you have an IMDB account, there is a page that shows you gender demographics and age demographics. IMDB gets these demographics from their IMDB members who have registered with their gender and age. Excel Data  | Annotated R Code | Disney version | Pixar version   Highlights and Considerations: •           IMDb ratings are rated on a 10-point scale. The average IMDb rating (across the 26,000 most rated movies) is 6.92 with a standard deviation of 0.96. This means most average movie ratings (68%) across   movies are within 1 point of each other. Therefore, while a decimal point difference between gender ratings might seem negligible on a 10-point scal...
Read more