There was an article in the local paper about the spread of COVID-19 in the county. They have a paywall, so you might or might not be able to read the article. It was posted several days ago, and Letters To The Editor are no longer A Thing (being replaced by comment sections where everyone argues), so my thoughts about this article are going to have to go in my blog.
The gist of the story is that they claim that the number of cases of COVID-19 is growing the fastest in ZIP code 92154, an upside-down U shaped ZIP code right on the Mexican border. This is true, in one sense, because this is the ZIP code in San Diego County that has the most COVID-19 cases overall. You need to have, on average, the highest rate of cases per day in order to end up with the highest number of cases.
The paper notes:
To determine average growth rates, The San Diego Union-Tribune first calculated a change in cases between each day from March 31 through May 31, and then averaged these totals for an overall daily growth rate.
Assuming that I am understanding what they mean here, this is equivalent to taking the number of cases on May 31, subtracting the number of cases on March 31 and then dividing by 61. They are not taking into account any changes in the rate of growth over the course of the pandemic. Also, they are assuming that the growth rate is linear, which is fine because since the middle of April it has been fairly linear. I would have been happier if they had calculated growth rates over shorter time intervals. I haven’t yet built out the analysis to look at growth rates over time, but the data that I’ve seen suggests that a naive model of “92154 is growing the fastest” does not tell the whole story. Once I run all the data, I’ll have a better story to tell, but mine would probably be “Things were pretty bad near the border, especially in San Ysidro, but all the other places that you think of as ‘poor’ or ‘Mexican’ or ‘immigrants’ are starting to catch up pretty quickly.” Not as catchy of a hot take, I know. I’m planning on changing my “new cases GIF” to show a moving average rather than the daily data, so we’ll see what that shows.
One thing that the newspaper’s analysis entirely ignores is population. The population of 92154 is 81,645. On the day that the paper did its reporting (June 3), there were 707 reported cases in that ZIP code. This meant that 86.6 per 10,000 people in the ZIP code had contracted the disease. However, this ZIP code surrounds San Ysidro (92173), which had 339 cases out of a population of 27,741, or a rate of 122.20 cases per 10,000 population. You can see this by going to my COVID-19 tracking site and setting the map to June 3.
By neglecting how the rate of growth changed over time (what we in the math biz would call the second derivative) as well as the relative populations, the newspaper article misses the fact that both San Ysidro and its surrounding ZIP code both had roughly the same number of cases per capita in early May (in fact, San Ysidro had a slightly lower rate) but that San Ysidro now has significantly more cases per capita. Based on this information, I would say that COVID-19 is spreading faster in San Ysidro than it is in nearby areas.
I know that the local newspaper can’t get too sophisticated in terms of the math that it uses in articles for the general public, but I would have liked for them to acknowledge that the growth rate changes over time and that the impact on a ZIP code depends both on the number of cases as well as the population.
Since I don’t do much, I don’t have much to write about. You might have heard that AI will advance to such a point that robots will steal our jobs. So I asked an AI system to write a blog post for me. This is what it came up with.
I finished the knitting project that I’d been working on. And it’s ready for blocking! It’s taken me a really long time to knit this, mostly because I’m usually too busy with deadlines to work on non-profitable knitting projects. It was fun to do something a little different, though. :)
I’ve set up one of my cameras exclusively for taking cat photos. Before, I couldn’t get any decent pictures of him. Today I have at least eight great ones. He’s not the brightest bulb in the cage, you see, so I need the flash to freeze him. When the camera comes near him, he freezes and opens his mouth. His expression says he’s saying something along the lines of “Hey, don’t you know you’ve been caught?” I’m a little sad. We had great hopes for at least one of our cats,
I started sewing another mask. I own a lot of homemade masks for someone who never goes outside. Most of them are the foam and latex kind of masks. Some of them have eye and nose holes but they have little in the way of ventilation. I added some ventilation holes to my latest version but it is still too heavy to wear in public. After all of the previous attempts I came to the conclusion that the only option left to me is the simple surgical mask. With its big eye and nose holes and small chin strap it provides me with a full face mask that is light enough
I’m taking a break from working on web page with the COVID data. I want to post another video that you all may find very funny. #thesearchcontinues #johnlennon #thebeatles #peace The Beatle Bible: The Fab Four’s Guide to the Good Life A good read if you want to learn about the Beatles. Many different viewpoints are given in this book. The Beatle Bible is written as a biographical yet informative look at their history, sound and content. The
- Our students do online homework and then they spend a lot of time kvetching about points in the bug reports. They seem to believe that they already own all the points and then the problems steal the points from the students; students will get a problem wrong due to what the student considers a minor oversight and then write in “GIVE ME BACK MY POINTS.” In a work-related conversation yesterday, I had to explain to a colleague that there is a good reason why I am not the Jay Powell of points. I think that in the students’ minds the problems are some sort of wealth tax on points and solving the homework problems is the equivalent of finding the correct tax loopholes.
- Every time a student in one of our programming classes submits a bad bug report (“The thing is not working.”), I hope that karma is keeping score. I imagine this student’s future self having to deal with bug reports just as bad.
- Bug report with TMI. Starts fine; goes dark. I’m fuzzing the details quite a bit,
but the bug report went something like:
I don’t remember me doing that many problems . oooooooooohhhhhhhhh I remember me doing those problems !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! I love math ! it is so fun !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! I mean ssssssssoooooooooooooo muchhhhhhhhhhhhhhhhhhhhhhhh ffffffffuuuuunnnnnnnnnnnnnnnnnn !!!!!!!!!!!!!!!!!!!!!! my house burned down in August 2019 so my family is living with my grandma . she is really sick !!!!!!!!!!!!!!!!!! please write back very soon!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! math is so fun !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! I love writing bug reports !!!!!!!!!!!!!!!!!!!!!!! it is so fun writing bug reports !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! I just love writing … …and reading!!!!!!!!!!!!!!!!!!!!!! it is fun writing bug
We’ll see if this GIF stays animated once I put it on the web. It might depend on your browser and the whims of ImageMagick. It’s animated on Facebook, which is something.
This GIF should be showing you the spread of COVID-19 through San Diego County, based on the total denisity of people infected (number of total cases per 10,000 population) by ZIP code. Orange-based colors are for ZIP codes with more than 10,000 people and more than 5 cases. Other ZIP codes are in gray-scale. Darker colors means a greater density of cases. Yes, this is total cases (not new cases). Yes, I might work on the GIF for new cases at some point, but this GIF was already kind of annoying to make.
Instead of devoting my recreational screen time to writing blog posts or watching TV or browsing knitting patterns that I might never make, I have been putting the infrastructure in place to make GIFs like the one I linked above. (But if it is not animating itself, I hope that the next version of it will stay animated.)
You might think: “Take some images, toss them in a directory, and then run ImageMagick over them all to get a GIF.” If you don’t know ImageMagick, it is the best ever command line tool for making images do things without having to open up some GUI-based software and click on things. You can script all sorts of amazing things with images.
Let me tell you about what it has taken me to create the images. I guess we’ll work backwards.
The PHP script is a bit more than I’ve mentioned above. What it does is it checks the MySQL database to find the timestamp of the most recent update that it has from the County. If that timestamp is long enough ago that the County should have published more data by now, it makes an API call (this is a very annoying API, btw) to get the new data and processes it to save in the database. The PHP script can use all of the data (both new and old) in order to create the JSON object. Note that the County’s API limits the amount of information that it will send you in a single call so you can’t get all of the data over all time by sending one query to the API. You either need to do a bunch of calls (SLOW!) and string them all together, or else you need to cache the older data so that you only need to get a little bit from the County. I went with caching because this data is only updated once a day, so it seems silly to keep looking for it in The Cloud if you know that there won’t be any new data for many hours.
(Since I’m doing all this on a server that a friend lent me, not some sort of professional hosting, I had to install the webserver, the database server, etc. and configure them all to play nice together.)
Now I just need to put the finishing touches on the bash script and set it to run on a cron so that I can have a fresh GIF each morning. And this is why I have not been blogging. Once the code is nice enough that I’m not embarrassed by it, I might make the GitHub repo public so that other people can make their own GIFs in a similar way.
Just to be clear, this is not a film that already exists. This is my terrible pitch for something that would be a terrible idea. Also that no one would allow to happen for so very many reasons.
There are a lot of people who do research on Kids On The Internet, so the fact that this idea would never happen is not going to keep you from hearing about kids on the internet. Someone I know just got a grant to study something along these lines. Tweens on the internet during social distancing? Something like that.
Roughly half a million accounts have been created on our site, and a lot of them belong to people who are currently kids on the internet. I don’t know how many of these accounts are still active or how many of them are still kids, but, oh my goodness, there are a lot of kids spending a lot of time on our site. Now, don’t get me wrong, there are also a lot of adults who spend too much time on our site as well. The thing about our site is that you kind of assume that everyone is a 13-year-old boy, but you can’t really know that.
So here is the pitch: pick some number of sub-groups of our users and watch their interactions on our site and then also meet them in real life. I think that this sort of story-telling works best with an odd number of subjects, so to make this imaginary pitch be feature-length, I’m going to say that we should look at five groups.
We’d start with a group that is somewhat unsurprising. A bunch of kids who post about their math contest goals and aspirations. Kids at the high school level. The sort of kids who aspire to be the sort of kids who would be featured in a film like Hard Problems. Send a crew to the kids’ houses, look at their wide array of math books. See the kids put together their schedules of how much time that they’ll spend studying each day. Talk to the parents about what they think of the kids’ math contest goals. Have the kids read some of their posts out loud. There are certain types of posts that they make all the time: “This Is My Math Contest Journey” or “Tell Me How Many Hours I Should Study Which Topics In Order To Win” or “How Many Points Do I Need To Score To Be Guaranteed Admission To Harvard.” As counter-point to this part of the narrative, we’d also talk to some of the long-term moderators who tell each new crop of students that there is not a fixed recipe to become a math contest champion. You should sit back and enjoy the problems. This is what everyone expects when it comes to the journey of math olympians.
As is true of the genre, we should probably skip around back and forth among the groups that we are documenting. We shouldn’t just have the Math Contest Strivers taking up the entire first fifth of my imaginary film.
I guess from here we would move on to a particular social group. I’m not going to mention their group by name because they try to stay under the radar – not in a troublemaking way, just in a minding their own business kind of way. A lot of the students from The Group are into math contests and take classes on our site. But they also put a lot of time and effort into maintaining their group. The Group is run via a system of private forums (students can only join if invited + approved by someone who runs the private forum, kind of like a secret Facebook group). They have a system of elections in order to select moderators and approve new members. Just about everything that I’ve seen on The Group has been pretty wholesome. There are a lot of homeschool students in The Group. Members of The Group has arranged – with their parents’ permission – to meet up outside of our site. There will be overlap between The Group and the math contest strivers. A lot of the math contest strivers won’t know that The Group exists, though. This group has been around for several years, and it is kind of amazing that they have been able to keep things up and pass it down to the next group of kids.
Not sure which group to move our focus to next. If we have a lot of money for our imaginary documentary, we could look at a forum that is devoted to a particular country’s students. There are a lot of students from this country who participate on our site. The way that education is run in this country is that there are national exams in order to decide who is admitted to higher education, and the competition is cut-throat. So we have all these students who are discussing math together but who are also competing against each other. They are posting rumors about the exam process. Additionally, the culture of these students differs from my culture in many ways, so the sorts of things that they say and do seem quite unusual to me because I am not familiar at all with their way of looking at the world. It would be pretty interesting to see these students in their regular lives. I might learn something about WHY DO THEY DO THAT. How does math contest preparation in their country differ from math contest preparation in mine? What do their homelives look like compared to the math contest strivers and the homeschoolers from The Group? If we have enough money to travel to their country, we could find out.
Here is where it turns dark. Everything that I’ve mentioned so far has been on our message board system. Most of the forums that I’ve talked about (with the obvious exception of The Group) has been done on the public parts of our site. And it’s a persistent public part of our site. If you went to our site, you could search through the forums and find the sorts of things that I am talking about. If you made enough wholesome posts about Harry Potter or whatever, you might even get invited to join The Group.
There is a part of our site that has a rated game and a chat for spectators. This is a game where students compete against each other solving math problems as quickly as possible. The winner gains rating and the loser loses rating. The is a lot of commentary and trash-talking in the chat. There are a lot of accusations of cheating. This is reasonable because some of the most committed players of this game want to be at the top of the leaderboard, and they will go to extensive lengths to gain rating and to keep it. Bugs in the system have been carefully documented, and they know how to exploit them in order to disrupt the game. This is the math website equivalent of the high school kids who smoke behind the gym. Some of these kids are 12. But some of them are the math website equivalent of the people who graduated several years ago but who still smoke behind the high school gym with the high school kids. Maybe they’re the ones who buy the metaphorical cigarettes? Probably not a lot of crossover with the wholesome kids from The Group. Also not a lot of crossover with the International Students. This is kind of a thing unto itself. You might see some of the math contest strivers trying to practice their fast-draw problem-solving skills here, but a lot of them will get eaten alive by the lifers who have memorized (“memmed”) the answers to hundreds (perhaps thousands) of problems in the database.
But what are these kids like outside our site? Does their ruthless streak extend outside of this game that they obsess over? I can’t imagine them running kitten rescues. I know from the timestamps on the games that a lot of them stay up all night playing against each other. Maybe instead of comparing them to the smokers behind the gym I should have compared them to the old men who hussle chess in the park.
Where do we go from here? I know that I’m pretending that we will cut back and forth between the groups and that we won’t be treating them in an isolated way. But that is hard because the International Students do keep to themselves to some extent – and are pushed away by the math contest strivers. The strivers might see the international students post about some non-US contest, and the strivers will insist that this should not be discussed in the main forums but rather in the forum specific to the country that runs the contest. But we should also end with something that ties everything together and on a somewhat positive note. We don’t want to delve into the niche petty arguments that a group of adults who should know better are making about each others’ approach to posing and solving problems about inequalities.
Maybe we go back to the beginning and look at the youngest kids who are just getting started with Middle School math contests. They don’t have the focus and dedication of the older and more competitive students. Some of them play the games on the site. Some of them are focused on math. Some of them spend a lot of time chatting. After we meet the next generation of students, we could speculate about where they will end up.
Once again we have two posts in two days. We’ll see how that goes.
Now that we are all working remotely, we are having a lot of telecons. Did you know that everyone’s computer speakers and microphone are terrible? They are! Also, my work computer is a Mac Mini, and while it has speakers (terribly), it has no microphone whatsoever, so I can not call in to meetings with this computer. Work bought me a headset, and it was quite a debacle of delayed shipping, the wrong version of the headset (USB-A vs. USB-C), and then buying the adapter needed to get it to work with my computer. I could say more about this, but does anyone want to hear? Probably not.
When I was shopping for a headset, I knew that I wanted to get a wired headset because I am really, really, really, really, really bad at paying attention. While waiting for this headset to arrive, I have been calling in to meetings on my phone. Since my phone is not tethered to my computer, during several meetings I have muted my phone and then sat on my couch to pet a cat, de-linted a sweater, and done other non-work things. Being forced to stay near my computer will help me pay attention. The wired headsets come in one-ear and two-ear models. I wanted to get a one-ear model. It was pretty much impossible to determine whether the one-ear models could be used on either ear or if they were designed asymmetrically. This is probably not a big deal for most people, but it would matter a lot of people with asymmetric hearing loss. And this is not some sort of no-name-brand headset. This is a Plantronics headset. The people at Plantronics should know that they are doing and should make this sort of information clear in their product descriptions.
But, get this, another feature of my headset is that if you download the controller software and use the headset through USB, it can adjust your noise exposure over the day. Based on the number of hours that you spend on calls, it will automatically adjust the volume so that your noise exposure stays within OSHA guidelines. It also has an anti-startle feature to mediate sudden loud sounds, and it has a maximum volume!
Guess what the software does not have? There are no controls to change the relative volumes of various frequencies. Since the headset software is already applying audio processing algorithms to the sound, this seems like the natural add-on. The microphones on hearing aids are not designed to work with headsets. Having a headset that can boost the missing frequencies means that people with hearing loss do not need to up the overall volume, which can damage their residual hearing.
Also related, a colleague confirms that the Google Meet auto-captions work best when the speaker is using a real microphone (like a headset, but a phone is OK) and worst when the speaker is relying on the computer microphone.
I told you not to expect more posts on a regular basis.
I’m full of excuses here. I couldn’t do any chores today because the water went out in the apartment for most of the day. You might think, “Wasn’t there a very disruptive plumbing project for pretty much the entire month of February so that this would not happen any more?” Yes, yes there was. I do not know the full extent of the situation, but it seems to be that there are a lot of parts of plumbing systems and a lot of things that can go wrong with them. This used to be a bad neighborhood before it was a trendy neighborhood, and this building was built during the “bad” era.
And I could not work on my hobbies because I can not motivate myself to knit a gauge swatch. I’m hoping to make a sweater with a picture of my cat on it. It would be so cool to have a sweater with a picture of my cat! But it might turn out badly in so many ways, and I am just not prepared for that sort of disappointment.
Also, it is forecast to be summer degrees outside at some point this week or next week. Probably not the right time to be wearing a cat sweater. And do I have enough yarn to make this sweater? I mean, I have many, many, many miles of yarn, but do I have enough yarn that is all the same type and all the same color to make something? Well, if the thing is purple, maybe.
And for someone who never goes outside, I have quite a surplus of homemade masks. Including two made with purple unicorn fabric.
People have a lot to say about the current global pandemic. My nerd friends like to talk about numbers relating to the current global pandemic. The County health department, bless their hearts, continues to make me anxious about their ability to work with quantitative information when they produce graphs like this.
For a while, before the County was producing any graphs, I was posting a graph every day on my Instagram story with the log-linear plot of total cases in San Diego County as well as the doubling rate, assuming an exponential model.
Allow me to make a nerd aside here, but a lot of people are calculating doubling rates in ways that make me very, very nervous. They assume some sort of geometric progression, find the ratio for the past few days, maybe average those together, do a bit of witchcraft with the log of 2, and BAM! they have a doubling rate.
If you are assuming that the number of cases is following an exponential model (which, around here isn’t the best assumption; piecewise linear is a better bet), do keep in mind that the case count is count data, which means that your model should assume a Poisson distribution and not a Gaussian distribution. So if you are using software like R, you would fit a generalized linear model with something like
mymodel <- glm(mydata$cases ~ mydata$days, family=poisson(link="log")). And then you could calculate your doubling rate with
log(2)/mymodel$coefficientsand the confidence interval with
log(2)/(confint(mymodelg2)[2,1]. For example, based on the data from the past week, we would say that the case doubling rate for San Diego County is 18.7 days, with a 95% confidence interval of 16.01 - 22.46 days. If you don’t report a confidence interval, you are just making stuff up.
But now the County publishes graphs! And some of them are bad graphs. BUT GET THIS, the County also publishes an API. (Spoiler alert: Not the best API.)
One of my friends got me a server so that I can make stuff with the County’s data. My test-rate graph does not yet do the 14-day moving windows that theirs does, but I did spend a fair amount of this afternoon making graphs instead of using water.
And tomorrow I will go back to “work” (at home) writing software to display data on webpages. Of course the work data is displayed on work webpages using all sorts of very sophisticated technologies, so it takes far more time (and tears) to make them.
subscribe via RSS