 Hello, everyone. My name is Kristen Pantagani. I am a MD-Ph to student at Baylor College of Medicine in my very last year of the program. And I also run the blog, YouCanKnowThings.com, which is a science communication blog that helps debunk misinformation and explain confusing things about the pandemic. And part of how I do that is using animated data visualizations. So this talk is about using GigiAnimate for science communication during the pandemic. So this was the very first plot I ever made on the pandemic back in March 2020. I remember just sitting in my office and looking at the cases in the US and kind of comparing how they were tracking with Italy. And I made the simple Excel plot and ended up sharing it on Facebook and it got a lot of engagement and a lot of people are really appreciative of having someone explain the data to them. And so I started sharing more and more of these and then decided to also make them a little bit fancier. So this is a plot I created recently using R and GigiAnimate looking at the number of COVID cases in the US and also the total vaccinated. And so I found that having these plots be animated has added higher levels of communication that I can't get an established plot. So that's what I'll be telling you about today. So GigiAnimate adds higher levels of communication compared to just a static Gigi plot. So first you can use time to get context, especially when you're plotting things like COVID cases over time. You can rescale the axes and this helps drive home things like exponential growth even more for people who aren't used to looking at data. You can use animated labels to make it easier to understand and capture people's attention in the half second they're going to think about your post when they're scrolling through Facebook or Twitter. And sometimes you just need a third variable and you can do things with GigiAnimate that you just can't do in a 2D static plot. And then finally having data move just grabs attention when you're sharing it online, which definitely helps. All right, so I'm going to take you through one of the scripts that I wrote that plots COVID cases and deaths over time. And so you can grab the scripts on GitHub. This QR code will take you straight there. All right, so this is the plot. So it's just in this case COVID cases over time and Harris, Texas, sorry, Harris County, Texas, which is Houston, Texas. The script lets you change the county so you can plot any county in the US. And so this is the static version of the plot and I'm going to take you through how we would animate it. So first, let's add time. So to do that, here's a simplified version of the plot. And so we have just our simple Gigi plot our data x axis is date y axis in this case would be COVID new COVID cases. And then to animate it all we do is take this Gigi plot and we add transition time and tell it what variable we want to animate over. So we're just going to animate over date in this in this example. And then we have to add this shadow mark which tells the plot to keep all the data points on the plot. Otherwise it'll only plot one data at a time and then they'll just disappear when it moves on to the next one. And Gigi animate has a really nice Gigi animate.com has a really nice tutorial to get more details about about the different options for these. All right, so this is what it looks like if we add the time variable. I like this because when the COVID cases shoot up you actually feel it because it's now animated over time. And so it gives a stronger sense of the rapid change of exponential growth. So the next thing we can do is have the axes we scale as as the plot is progressing. And to do that we just add this simple view follow to our animation or animation code. So you can see that the y-axis and the x-axis are restailing every single time a data point is added to the plot. And this helps even more drive home things like exponential growth because you see not only the plots shooting up, I'm sorry the points shooting up, but you see the axis is having to rapidly change to accommodate for the rapid change in COVID cases. So it's just a more dramatic visual of rapid change over time. All right, so this is another plot I made of different data. So this is weekly deaths from pneumonia and influenza and then recently COVID over the last several years. So I plotted this just to show, you know, kind of what the baseline is for the for seasonal flu and then watching just a second you'll see COVID. So then it shoots up. And so having this axis rescale really makes that change really dramatic and drive some of the point that the COVID pandemic, the amount of deaths that we are seeing is way different than previous flu cases. And all these animations you can see on my website. I'm sorry I don't have time to let them play through multiple times, but if you want to go look at them again, they're on my website. All right, so next you can add animated labels to make make the plot easy to understand. So if you're just scrolling through your Facebook feed and you see something like this. It's interesting because it's moving but it doesn't immediately jump out at you what the plot is. And this is the plot we just we just made of our COVID cases in Houston. And I will tell you people don't really read access labels. If it's going to be on Facebook or Twitter, you can't just rely on the access label. So to do that, what we can do is add an animated label and do that with GM text repel we add it to our our gg pot. And we use we tell it what label we want. So in this case I'm using a variable from my data frame as the label you can also just put a string in here. And then we still animate it and one thing you have to do is your shadow mark, you have to exclude layer two in this case layer two is this GM text repel layer of your gg pot. Otherwise your pot will just be entirely full of labels and global. So now this is what it looks like we have this label of new cases that's tracking along every time one of our data points is plotted. And this is now what grabs your attention. So if you're looking at this on social media, you immediately know what it's talking about it's plotting new cases. And because we're in a pandemic clearly talking about new cases. And so this is a way for people who you know aren't used to looking at data or people who are just very distracted because they're on Facebook to really like instantly tell them what your data is about and grab their attention. You can also use GM text repel to annotate other things besides just what the data is so in this case this is the measles cases in the US before and after the first measles vaccine. And so here I'm just animating a single data point, I'm sorry labeling a single data point of when the vaccine was introduced. And last, sometimes you just need a third variable so there are things that you just can't plot in a static 2d plot that dg animate really allows you to plot. And here's an example of that. So this is showing the confusing mathematical concept that as the percent of a population becomes increasingly more vaccinated than the percent of all the hospitalizations for covid were vaccinated people will also increase. And so in this case, the data were or the variable we're animating over is not the x axis and it's not the y axis. It's a totally separate variable in my data point. But it allows us to kind of show three dimensions of the data because of the added dimension of time. Alright, and here's another example of that. So this is looking at just visualizing the difference in different are not so on the left we have are not 2.5 on the right are not six. And the thing where the variable we're animating over is not x or y. It's a totally separate period. Alright, and lastly, animation helps grab attention. So the goal is to communicate with the public, and it certainly helps to have more eyeballs on the graph. And animation helps helps you do that because people like things that move online. And so these are two of my most successful animations if you define success as number of retweets and likes. And I will certainly say that not everything gets this level of engagement. I plotted many that don't get nearly this much. But it does go to show that having that animation part in there really helps just engage people and then the more engagement you have the more shares you have and then the more people are seeing your message. All right, that's all I have for you, you can find me on Twitter at camp and the ghani or on Instagram that you can do things. I also as I mentioned have many of these visualizations on my blog that you can do things calm so if you didn't quite catch something in them and want to go look again you can find them there. And here again is the code for the animation I walked you through. Thank you.