Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Display of active cases vs time #84

Open
nikohansen opened this issue Mar 25, 2020 · 9 comments
Open

Display of active cases vs time #84

nikohansen opened this issue Mar 25, 2020 · 9 comments

Comments

@nikohansen
Copy link

Thanks for sharing this, very nice project! I'd like to make a few small suggestions on the time evolution graphs:

  • it seems that the active cases over time should be normalized by the number people (country size). I found the closed issue Normalize time evolution #3, but it seems not to be effective yet? The number of cases in itself is a meaningless object, in particular when compared across countries. The effect of a single case is always relative to the number of health care professionals who can take care of it.

  • Using log as default scale: if I click a new country, the scale switches back to linear and displaying these data on linear scale is to me close to useless. Of course, it is good enough if the scale does not switch back.

  • I'd prefer if the two figures, world map and time evolution, were on top of each other rather than side-by-side, such that they can be displayed larger on the screen.

I made a similar effort that you can find here

with my rationales given here but I'd also be happy to join forces. I did it because I couldn't find a single source plotting new cases per day on a log scale for a variety of countries.

@poldrack
Copy link
Contributor

I agree that either log should be the default, or the scale choice should be sticky when one changes the country selection

@GaelVaroquaux
Copy link
Contributor

We have now implemented the normalization for the time series.

With regards to the log vs linear scale: I agree that the resetting is a bug (it's dash / javascript, and I do not know how to fix it). However, given that our target audience is the layperson, I would rather have linear scale by default: people do not have intuitions about exponential growth (including decision makers). I think that the best would be to fix the bug that reset. Such a contribution would be incredibly helpful.

With regards to you last point: try blowing up the fonts for the page (ctrl-+ under firefox): the layout will adapt.

Cheers!

@nikohansen
Copy link
Author

nikohansen commented Mar 25, 2020

try blowing up the fonts for the page (ctrl-+ under firefox): the layout will adapt.

I (obviously) tried. This is what I get in Chrome:

Screen Shot 2020-03-25 at 20 29 42

The figure I am interested in now occupies a far too small amount of my screen, to my taste. On the positive side, the legend is readable ;-) If I make it larger, the legend flips in the next row and becomes huge and difficult to access. On the positive side, then the figure occupies the full screen.

@GaelVaroquaux
Copy link
Contributor

Thanks for the screenshot. You may be hitting a rendering bug that makes the table too large.

I suspect that you are on a high-DPI screen. Can you tell me your DPI, as detected by your browser (visible be visiting the following page https://www.infobyip.com/detectmonitordpi.php ), your screen width, and the br

@nikohansen
Copy link
Author

Rightly so suspected. The page says 192 x 192 (on a recent 15.4" 2880 x 1800 Macbook pro).

@GaelVaroquaux
Copy link
Contributor

OK, I'm working on addressing this in #88

(you can check out the link to circle CI artifacts, in the status badges of the PR, to visualize a rendered webpage).

@nikohansen
Copy link
Author

nikohansen commented Mar 25, 2020

Fixed the font size in the legend table, but the table width is unchanged (i.e. its widths is not adapted to the font size).

@GaelVaroquaux
Copy link
Contributor

How about the version of #88 ?

@nikohansen
Copy link
Author

nikohansen commented Mar 25, 2020

No fix of width with a101234, as far as I can tell (sorry, was distracted).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants