Charts and Graphics
-
At the risk of turning this into a political discussion...(and yeah, I know, if I'm already feeling compelled to put in this disclaimer, it probably already is). But if we can get even a semi-coherent discussion going, I'm willing to risk getting this locked down/deleted/whatever and I'll refrain from further voicing my opinion on this matter. With this out of the way... There's been so many discussions about discrepancies between in the ways different states or even entire countries test and report infection rates and hospitalizations and the fact that some people are completely asymptomatic and false positives and such, all numbers are questionable when you're comparing apples and oranges. As developers, we should all understand (I hope) that when you're working with data, all of your data points have to be measuring the same thing, otherwise, you end up with garbage in, garbage out. This is how one can make the claim that "the more people you test, the more infections you're going to detect, duh!". People can skew, twist and spin numbers to fit whatever agenda they're trying to push. I'm absolutely NOT denying there's a pandemic out there and people are dying from it. But what I'd like to see is a graph, over time, showing the total number of deaths--IRRELEVANT OF THE CAUSE--plotted year-over-year. This
many people to die in any given month of any given year, then I want to see those figures for a number of years.Then lets compare this baseline with 2020. IMO, how much 2020 deviates from that baseline (x+n) would be a much better indicator of the pandemic's cost in human life. Yes, that would include deaths related to the increase in stress, depression, suicides, people avoiding the hospital out of fear of getting infected, etc. But isn't that the figure we should be trying to find out anyway?These are reasonable comments we need to ask about any data we get. The comments are not that politically explosive. Still I was hoping to avoid them in this thread. I want to focus technically on the way the data is presented complete with animation. How many things is shows in a short time. Things you would never see by looking at tables of numbers. The data is grim no mater the details of how one comes by it. Just for a fraction of second ignore the data and look at the technical art.
So many years of programming I have forgotten more languages than I know.
-
At the risk of turning this into a political discussion...(and yeah, I know, if I'm already feeling compelled to put in this disclaimer, it probably already is). But if we can get even a semi-coherent discussion going, I'm willing to risk getting this locked down/deleted/whatever and I'll refrain from further voicing my opinion on this matter. With this out of the way... There's been so many discussions about discrepancies between in the ways different states or even entire countries test and report infection rates and hospitalizations and the fact that some people are completely asymptomatic and false positives and such, all numbers are questionable when you're comparing apples and oranges. As developers, we should all understand (I hope) that when you're working with data, all of your data points have to be measuring the same thing, otherwise, you end up with garbage in, garbage out. This is how one can make the claim that "the more people you test, the more infections you're going to detect, duh!". People can skew, twist and spin numbers to fit whatever agenda they're trying to push. I'm absolutely NOT denying there's a pandemic out there and people are dying from it. But what I'd like to see is a graph, over time, showing the total number of deaths--IRRELEVANT OF THE CAUSE--plotted year-over-year. This
many people to die in any given month of any given year, then I want to see those figures for a number of years.Then lets compare this baseline with 2020. IMO, how much 2020 deviates from that baseline (x+n) would be a much better indicator of the pandemic's cost in human life. Yes, that would include deaths related to the increase in stress, depression, suicides, people avoiding the hospital out of fear of getting infected, etc. But isn't that the figure we should be trying to find out anyway?Some student at John's Hopkins did an analysis with the CDC's data. It got removed, but is still available in PDF. It showed no excess deaths. I read it, but didn't check the work, myself. Also, Dr Clare Craig (@clarecraigpath on twitter) does a lot of this type of analysis.
-
Some student at John's Hopkins did an analysis with the CDC's data. It got removed, but is still available in PDF. It showed no excess deaths. I read it, but didn't check the work, myself. Also, Dr Clare Craig (@clarecraigpath on twitter) does a lot of this type of analysis.
Some "student" ... Chinese funeral urn sales have apparently sky rocketed. Urn speculating?
It was only in wine that he laid down no limit for himself, but he did not allow himself to be confused by it. ― Confucian Analects: Rules of Confucius about his food
-
At the risk of turning this into a political discussion...(and yeah, I know, if I'm already feeling compelled to put in this disclaimer, it probably already is). But if we can get even a semi-coherent discussion going, I'm willing to risk getting this locked down/deleted/whatever and I'll refrain from further voicing my opinion on this matter. With this out of the way... There's been so many discussions about discrepancies between in the ways different states or even entire countries test and report infection rates and hospitalizations and the fact that some people are completely asymptomatic and false positives and such, all numbers are questionable when you're comparing apples and oranges. As developers, we should all understand (I hope) that when you're working with data, all of your data points have to be measuring the same thing, otherwise, you end up with garbage in, garbage out. This is how one can make the claim that "the more people you test, the more infections you're going to detect, duh!". People can skew, twist and spin numbers to fit whatever agenda they're trying to push. I'm absolutely NOT denying there's a pandemic out there and people are dying from it. But what I'd like to see is a graph, over time, showing the total number of deaths--IRRELEVANT OF THE CAUSE--plotted year-over-year. This
many people to die in any given month of any given year, then I want to see those figures for a number of years.Then lets compare this baseline with 2020. IMO, how much 2020 deviates from that baseline (x+n) would be a much better indicator of the pandemic's cost in human life. Yes, that would include deaths related to the increase in stress, depression, suicides, people avoiding the hospital out of fear of getting infected, etc. But isn't that the figure we should be trying to find out anyway? -
Some "student" ... Chinese funeral urn sales have apparently sky rocketed. Urn speculating?
It was only in wine that he laid down no limit for himself, but he did not allow himself to be confused by it. ― Confucian Analects: Rules of Confucius about his food
-
Some "student" ... Chinese funeral urn sales have apparently sky rocketed. Urn speculating?
It was only in wine that he laid down no limit for himself, but he did not allow himself to be confused by it. ― Confucian Analects: Rules of Confucius about his food
-
These are reasonable comments we need to ask about any data we get. The comments are not that politically explosive. Still I was hoping to avoid them in this thread. I want to focus technically on the way the data is presented complete with animation. How many things is shows in a short time. Things you would never see by looking at tables of numbers. The data is grim no mater the details of how one comes by it. Just for a fraction of second ignore the data and look at the technical art.
So many years of programming I have forgotten more languages than I know.
-
Some student at John's Hopkins did an analysis with the CDC's data. It got removed, but is still available in PDF. It showed no excess deaths. I read it, but didn't check the work, myself. Also, Dr Clare Craig (@clarecraigpath on twitter) does a lot of this type of analysis.
This thread was started not to talk about the data but the way the data is presented. Ignore the data for just a split second and look at the art of presentation.
So many years of programming I have forgotten more languages than I know.
-
michaelbarb wrote:
every so many pixels high represents
Oh, I typed in Georgia where it says "Search for your state" and there it had a line that showed 0 until October.
michaelbarb wrote:
how much data was presented to you in such a short time.
Ya, it's OK I guess. I have done lots with charts and this is just another one in the long list of charts.
-
you know it shows deaths per 100,000, rounded (or possibly truncated) to an integer? See the blips in April & May.