 It's LinkedIn Learning author Monica Wahee with today's data science makeover. Watch while Monica explains how she curated plot options you can configure for a complex plot in R. Hello everyone. If you follow my blog, you may have noticed I made a pretty explicit post about how to use the UPSET R package to make the most delicious UPSET plot. This visual you see here, this diagram, is from that blog post. I'll link you to it in the description to this video. Okay, so if this is the first time you have seen this image, you are probably overwhelmed by it. I admit it is a pretty busy image, but I encourage you to actually read the blog post. If you do that, then you might see this image differently. In fact, this image might inspire clarity. A lot of my learners want to learn more about data curation, so I thought I'd make this video and blog post giving you the backstory behind this image. So, why did I make this data curation image? Because I had a confusing data problem I wanted to solve. Here was my problem. What I had to do was assemble a huge vector in R that consisted of many different text options. See this comment code on my blog post? This was my first attempt at figuring out this vector I had to make. You will see what I did was write the name of the option from the UPSET R documentation, such as intersection size title. And then I wrote what I called it on my plot example, which was counts of patterns of conditions. I actually copied something like this from a comment I found when looking for more information on this set of options on the internet. But I was still too confused when looking at this to know what to actually put in the vector and know how to troubleshoot it if I wanted to change how the resulting plot looked. Here is something I could have done. I could have made a slide like this. See that? I have an example of some code that makes a vector called text underscore scale underscore options three. Then on the right side of the code I have an example of a set of numeric text options and in red I've written the name of the object in the plot to which they are referring. But this didn't solve the problem and I'll show you why. Okay, the first thing that is supposed to go in that vector is the text scale option for an object referred to in the documentation as intersection size title. So where is this intersection size title? I mean maybe it makes sense to everyone but me, but I couldn't tell which of these titles was that and that led me to here. This finally helped me. I used an example plot which was my example plot the one I was trying to configure and I annotated it with these ugly red labels. Let me show you how I did it. Okay, here we are in PowerPoint. See this plot image? I'll click on it. I exported this image from R and I just inserted it into this PowerPoint slide. Now see these ugly courier new type font labels that are this alarming red color? These are just text boxes I placed on the image. And this is just an arrow shape. See it's kind of like an art project. I made the arrow and the text the same color. So they look like annotations. You know like if you try to put something together that you buy from IKEA and you have to look at the diagram. It has these labels and arrows all over it. Oh and here's another example. See this annotated vector? I showed you this diagram before. These are just all text boxes. These ones here that are tilted. See this rotation thing here? I just rotated this. I used aerial narrow so it would be less busy. Yeah I fussed with the font, color, size, and location a lot. I fussed with arrows a lot in general. You have to play around with the image until it really speaks to you. Until you go away and come back and you think it's still clear to you. So as you can see even for someone who does curation all the time like me, this was kind of a challenge. Here are a few tips I would give to someone who is just beginning to build their skills in data curation. First, don't be afraid to try different things until you get a visualization that speaks to you. Like I did. Save all your work too. You might find yourself making amalgams and collages of it as part of trying different things. Early on in your data curation, consider making a table. Now this would not have worked in this case due to confusion over naming. But you will see if you hang around my YouTube channel enough that I make a lot of crosswalk tables as data curation files. My third point is that if you are making a curation file and you find it is getting too busy and too cluttered, the best thing to do is consider splitting the ideas into two files. You might be trying to document too many ideas in one image. I managed to squeeze the color and the text scale option documentation on one image, but it was pretty full. Probably anything additional would have needed its own curation file. And how much do you curate? It depends upon your audience. Remember when I showed you that annotated vector with all those diagonal labels? I didn't make that before because I was doing this project as a solo analyst. However, if I had been working with someone such as my forever intern Natasha, who I always talk about, I would have definitely created some sort of image like this. This is why Natasha and I tend to get along. If we get into some confusing argument about data, I just make some sort of diagram and PowerPoint and it's all good. Finally, my last curation tip is to leverage as many visual elements as possible to your advantage. For example, you will see I used Courier New as the font to annotate the text scale options. That's because it is ugly documentation font. Those of us in technology know it as a classic. We do not expect it as being part of the plot. The red color also jumps out. Font type, size, color and any images you use, they all communicate a lot. Remember, a picture is worth a thousand words. And that's the whole point of data curation. Communication. Thank you for watching. If you liked the video, then please give it a like. Also, I invite you to look around my channel. And if you like what you see, subscribe. I hope you are enjoying your day.