R for the Rest of Us Podcast Episode 2: Nassos Stylianou and Clara Guibourg

In 2017, BBC data journalist Nassos Stylianou was working with a backend developer on a particularly large data set. Nassos was primarily an Excel user at the time, but this data was too large for Excel. Seeing the developer work through the data with ease, a light bulb went off for Stylianou: if he and his data journalism team learned to use R, they could do this type of analysis on their own.

This realization began a journey into R. This journey, which started with needing to analyze data too large for Excel to handle, would ultimately end up in a very different place. In 2018, Stylianou, his colleague Clara Guibourg, and their team created a custom ggplot theme to create plots that match the BBC style. The code in the bbplot package is a great example of the value of developing a custom theme. But the real story of the creation of bbplot is not just about technical tools. Through learning R and creating a custom theme for others to use, Nassos, Clara and their colleagues would change the culture, remove bottlenecks, and allow the BBC to be more creative with their data viz.

To understand how big these changes were, it’s helpful to understand what things looked like at the BBC before bbplot. In the mid-2010s, journalists at the BBC who wanted to make data visualization had two choices:

They could use an internal tool. This tool could create data visualization, but only the predefined charts it had been designed to generate.
They could use Excel to create mockups and then work with a graphic designer to finalize the charts. This approach led to better results, and was way more flexible, but required extensive back-and-forth with a designer. As Stylianou described it, working with a designer “is just a very time-consuming workflow if you think of how many visualizations the BBC does.”

Neither of these choices was ideal. And this limited set of less-than-ideal choices led to a limited output of data viz.

That would all change when Stylianou, Guibourg, and their colleagues realized that R, the tool they had decided to learn for data analysis, could also do data visualization. As they began playing around with ggplot, they quickly saw its power. Guibourg said she found it “immediately addictive when I started working with ggplot to make charts.” No longer limited by the BBC’s inflexible internal tool, she found that ggplot was “completely flexible in a way that was just completely new to me.”

The biggest change, though, came from not having to work with a designer. Not because the designers were bad (they weren’t), but because ggplot allowed the BBC data journalists to explore different visualizations on their own. Working with a designer required the journalists to have a fully-formed idea that the designer could take and improve upon. Working in ggplot allowed BBC data journalists to explore different data viz ideas.

Clara Guibourg believes this freedom is what explains the addictive quality of ggplot. As she told me, “even before we got anywhere near having a production-ready chart, just trying things out, visualizing things for the first time” was completely captivating. Having learned the basics of ggplot, she saw that “you can make like the simplest chart with just a couple of lines of code.” Being able to explore different types of visualization on her own led Clara and others to produce more data viz than they had previously.

As the BBC data journalism team improved their ggplot skills, they realized that it might be possible produce for more than just exploratory data viz. They had learned to use R for data analysis and they were starting to use it for exploratory data visualization. Could they go all the way and create production-ready charts in R that could go straight onto the BBC website?

Stylianou, Guibourg, and their colleagues set about looking into what would be involved in creating production-ready charts from R. They realized that so much of this work involved small tweaks. What font should they use? Where should the legend go? Should axes have titles? Should charts have grid lines? These questions may seem small but they have a big impact. Having consistent answers to them is what enabled BBC designers to turn Excel mockups into high-quality data viz ready to go on the website. As the BBC data journalism team dug further into ggplot, they realized that they might be able to write code to make their data viz production-ready. They realized that, if making production-ready charts required asking question about fonts, legends, axes, and grid lines, ggplot had the answer. And the answer was to make a custom theme.

Nassos's website and Twitter
Clara's website and Twitter

Learn More

If you want to receive emails when we publish new podcast episodes, sign up for the R for the Rest of Us newsletter. And if you're ready to learn R, check out our courses.

Audio Version

Video Version

The video version has a walkthrough of how they developed the bbplot package to give the BBC a custom ggplot theme for all of their data viz work.

Resources Discussed

Read Chapter 3 of R Without Statistics, where this conversation is summarized
bbplot website
BBC Visual and Data Journalism cookbook for R graphics

Transcript

[00:00:00] David: Hi, I'm David Keyes and I run R for the rest of us. You may think of R as a tool for complex statistical analysis, but it's much more than that from data visualization to efficient reporting, to improving your workflow. R can do it all on this podcast. I talk with people about how they use R in unique and creative ways.

[00:00:18] Join me and learn how art can help you.

[00:00:20] I'm joined today by Clara Guibourg who is a multilingual data journalist working at Swedish public radio, but who used to work at the BBC news and by Nassos Stylianou a senior journalist at BBC news who specializes in data driven journalism there.

[00:00:37] Welcome to both of you and thanks for being with me today. I appreciate.

[00:00:42] Nassos: Yeah. Thank you. Very.

[00:00:43] So I'm excited to talk to you. I came across your work specifically because you released a package called BB plot, which makes it possible, which uses a custom theme within Gigi plot. To make plots take on the BBC style. So I'm excited to talk to you about that, but before we do that I'd like to know just a bit about your background specifically as it relates to R so maybe Clara, if you don't mind starting when did you start using R and what changed for you when you made that switch to R.

[00:01:12] Clara: Sure. I probably started using our exactly around the same time that I started working at the BBC. So previously I'd I've been aware of it for a couple of years and been meaning to teach myself. I'd been trying to teach myself and I'd been doing a couple of, Tutorials online and that kind of typical thing where you do a tutorial and you follow it perfectly and you think, oh great.

[00:01:30] I'm great at this. Now I can do anything. And then you tried to do even the simplest thing on your own and you can't get it and you just give up. But once I started at the BBC, I actually committed to more fully. And basically, so that would be start of 2018. So yeah I started then and basically never looked back

[00:01:45] And I think that I think that what helped me get started with it then was that there was just a really supportive environment at the BBC for learning internally. There was, I had really helpful colleagues and there was a lot of you really felt like you could take the time to ask questions and learn learn.

[00:02:01] David: And just outta curiosity, what tool or tools were you using primarily before you made that switch?

[00:02:05] Clara: So I was working, I was working previously in, in Excel, basically spreadsheets. And in terms of visualizations, I I would've been using chart tools,

[00:02:13] David: great. What about for you Nassos? When did you take up R and how did that change the work that you were doing?

[00:02:20] Nassos: Yeah, I think I probably started in a similar way to class using kind of Excel spreadsheets and for mapping sort of tools like Q, G I S to visualize things. And then at BBC, as we started doing more as the team. joined up with kind of data scientists and programmers, the data journalists, the more sort of editorial journalists side of it.

[00:02:44] We started seeing a bit wider world out there really. That was using our using Python, using these things and working really closely with a a developer was that the BBC at the time whose background was backend development, and he was using R for a lot of the data projects.

[00:03:00] We. Do so it would be any task that was too onerous for Excel would go to him. and he, as you can imagine, he was very popular within the team. And sitting next to him and getting a feel for the work he would do. I slowly started learning art yeah.

[00:03:15] Of him and stack overflow. Really? I would say and. I think it was a slow, a really it's a weird process. It was really quick at first it was oh, this is amazing. And then probably a few months of feeling I have literally made no progress. I'm doing the same things over and over again until you discover functions and you discover a few other things and then. like you look back and say, God, I, I can't imagine doing this like manually that I used to do. And then you realize, yeah I have progressed really rapidly, but I think it started off from my side. Very much analysis, I wanna do more involved and more and heavier. Working with bigger data sets use it for programming and analysis purposes.

[00:04:03] And the visualization came after that.

[00:04:05] David: Interesting. So when What was it that led you to realize, oh, we should use this, not just for analysis, but also for the visualization.

[00:04:18] Nassos: workflow issues, I guess it, it just we'd come to a stage where this might be the disadvantage in a way of a larger team. In that there are like there at the BBC, we had designers who would work on graphics and we would get a certain distance in our, and work it, like analyzing it and then exploring it using ggplot and using the sort of custom themes.

[00:04:40] And then we would sort. Work with a designer or someone who's a bit more specialist on the design side of things and export something for them. And then they would make it sing really and work on it. And, but that is just a very time consuming workflow. If you think how many kind of basic charts or how many visualizations you want to like the BBC does, especially another thing to bear in mind with the BBC is there's 40 odd world service languages as well, who also kind of reversion things.

[00:05:10] So you wouldn't really want Thinking back, it was workflow issues that were like, there has to be a better way, rather than us working in our doing the analysis, doing the visualizations as exploratory data visualizations, exporting as a PDF, then getting it to someone in illustrator who would then spend time cleaning that up.

[00:05:28] So it was that final bit of the workflow that we were what could this be done in our altogether? And it was more of an exploration and I guess that's around, around the time sort. Clark came in and we started properly looking into how far can we go with getting templates, making all this stuff that we are doing production ready?

[00:05:48] Just in our.

[00:05:51] David: Yeah. How many kind of folks were there on your team who were at that point? It sounds like you were both like getting into art at the same time. How many other folks were there, who were doing that as well?

[00:06:08] Nassos: trying to think back probably five around five of us. Isn't there. Yeah.

[00:06:12] Clara: Yeah, I think that sounds about right. I would say that most of the data team was already using R for analysis. So it was a question of a few of us being A handful of us. Yeah, probably that were curious about trying to trying to solve these problems that, that Nassos was talking about of trying to figure out a way of getting overcoming these like individual little obstacles that we had in order to get like production ready graphics from start to finish in R.

[00:06:37] David: Yeah, that makes sense. I guess I'm just curious because I don't, as an outsider, I hear BBC and I think, thousands of people but obviously it's gonna be a small team who's doing that, that one particular piece.

[00:06:48] Nassos: Yeah. I think there was a couple of people who knew her really well, but again, from the data science side, from the back end side nobody came into the team at that time. being a specialist in ggplot that then moved us on. It was we were all learning together and researching together and adding components together in like, how far can we go with this?

[00:07:10] It was a collaboration where we were all of a similar level really, but then each person might have find an answer to here's how we make the footer left align and really. very specific things like that were it's crazy how important they were at the time.

[00:07:27] Clara: Yeah, but.

[00:07:28] Nassos: it was like, if you're gonna generate like hundreds of graphics, you want the footer to always be the same proportion and you want the BBC logo, the blocks to always be in exactly the same place. It's yeah.

[00:07:39] Clara: I remember long discussions about the exact should the font be, should it be nine or 10 spaces above? Yeah, no, I definitely agree with that. But I think that was real. It was a really big benefit for us that it was just an an exploration where we were, like you say Nassos, that we were learning together and trying to trying things out.

[00:07:58] I think that kind of collaborative and experimental approach really was why I think that kind of brought our, like our, the group's level, like the, our knowledge forward faster than it would've done. Otherwise. I think if that makes sense, I.

[00:08:13] David: Yeah, definitely.

[00:08:14] How did you, you've talked about how you were all like exploring art at the beginning. How did you get from that point to the point where you realized, Hey, we could actually, take what we've learned and make a custom theme and create this package. Do you remember was there a light bulb moment or how did that come about?

[00:08:32] Clara: I don't think that there was a Correct me if you disagree Nassos, but I would say that it was such a it was just like constantly evolving thing. Like you started out. Point where where we were just start, firstly just playing around with Gigi plot trying out how can we get the right font for it to look like a BBC chart, and then the next obstacle, which was, how can we get the BBC logo to all of our charts? How can we export them with a little horizontal line that needs to go across.

[00:08:55] Nassos: Yeah.

[00:08:56] Clara: The future of all the charts things like that. And then once we'd developed like functions that worked for this, trying to work out, how do you actually create a package that we can, that was like the next step?

[00:09:05] Which I don't think that any of it was oh, this is like this one big light bulb moment where we brought it forward. It was just constantly making little fixes and taking the next step.

[00:09:14] Nassos: Yeah. Definitely I guess an important thing to bear in mind is that the starting point was not, how can we make a BBC chart package? The starting point was yeah, I think the starting point was literally, can we do make a production ready graphic that does not need to go. It does not need any modifications that can go straight from our, onto the BBC website.

[00:09:39] That was the starting point. And then I. No, I, yeah, I would agree. Sorry. It wasn't a light bulb moment, but the more of those we ended up doing and publishing on the site, the more we're like, Hey, what about, packaging this up and sharing even more widely outside of the five of us. And I guess even before the step where it's a package that you can download and use yourself, there was a stage where the kind of. Ambition, the thought was like, how can we actually get other people in the BBC using it? And then that aren't super experts in our role. Don't use it day to day and might know a little bit. So that was the other internal thing going on is how can we get other people outside of the team to use it?

[00:10:26] David: That makes sense. And so you talked a bit just now. About how creating this custom theme and putting it in a package allowed you to create graphics directly from our, and put them straight onto the BBC website. You touched on this a bit earlier, but I'm wondering if you can talk in a bit more depth about what that process was like prior to working prior to having the custom theme.

[00:10:49] You talked about how you worked with the design team to finish the graphs. So can you walk me through step by step so we can get the contrast of what it looked like prior to having this theme?

[00:10:59] Clara: Sure. Before we had this before we had this theme, before we were working with the gigg plug for visualization, we basically had two options for graphics. We could either use the there was an there or there is, I presume there is an in-house chart tool which you could use for quick turnaround things.

[00:11:14] Basical. It very easy to use, very rigid and, doesn't really offer any flexibility in terms of what you could you get your input, your data and you get a line chart that looks just so basically the other option was to would be to to order to work together with the designers on something which obviously means a, it is something that takes longer time.

[00:11:34] So it only really be used for longer term things.

[00:11:39] David: Yeah. And so that in that case you would make something in Excel I'm assuming, and then have to do a bunch of back and forth with the designer. Is that how it would typically work or was it. Process.

[00:11:49] Clara: yeah. Something, basically either. Yeah. Yeah,

[00:11:53] David: Okay.

[00:11:54] Nassos: Yeah. It was either yeah, I guess it was either export a PDF from our here's, how we wanted to look or an SVG with kind. Standard ggplot stuff, either something in Excel, either something in the chart tool saying here's the base of it. Can we add stuff onto that? Can we, add a line for a forecast, make the sort of side of it, half of it gray.

[00:12:14] Can we do this thing, which wasn't in the standard chart tool was a little harder. Can we add an annotation? Can we add an arrow pointing to so anything that was additional, which is it sounds, you've got a chart tool. You've got ways to do it, but I think a lot of these little additions, like an annotation are fundamental to storytelling really.

[00:12:35] So whilst they feel like, you know, extras for us, we of really wanted to get there from the, to have a way to get there. And I think what's also important in terms of kind of thinking of the wider BBC and the sort of design team is that it wasn't really a fun job for a designer to get a brief that was, can you recreate this or can you add something like this, as opposed to them really putting their creative skills to infographics, to UX, to design pages, to do much more creative stuff than slightly monotonous, reversioning a graphic that's almost there, but not quite from a data journalist. So it helped benefit both sides. It freed up and it has freed up designers to do much more creative and much more illustrative stuff, as opposed to recreating charts all day. I. Yeah.

[00:13:25] David: that's really interesting. I've never thought about that as a benefit in an organization like yours. And I don't know if for me being able to create everything in Gigi plot actually makes me more creative. I feel like, because I can, as opposed to having to sit down and think, okay, I want.

[00:13:40] This here, let's try this here, but then you get it back from the designer. I imagine there were times and you're like, oh, that isn't quite what I wanted. But then you feel bad if you keep asking them to redo it. Whereas if you can do it all in Gigi plot, you can explore, okay, let me try, making the gray slightly lighter, whatever the thing is, it probably allows you to do more creative things on your end because you don't have to constantly go back and forth

[00:14:04] Nassos: Yeah. It was exactly that. Yeah. I can't imagine that designers were I, no I'm sure that designers are much happier sense as opposed to us standing over their shoulder moving things around live. Whereas yeah, that is, 80% of making a chart GE plot, moving things around till you get it.

[00:14:20] Clara: Yeah, I was thinking about this, cuz I definitely found it like immediately addictive when I started working with ggplot to make charts, like even before we got anywhere near having a production ready chart, just trying things out, visualizing things for the first time.

[00:14:32] And I was trying to work out why. That was, and I think it is that, that you were saying that you can make like the simplest chart with just a couple of lines of code and you've got something that's just there. And then you realize that like the possibilities to tweak, this are limitless.

[00:14:47] Like I can do anything now. and you can go as in depth, as you want, just moving things, moving the tiniest things around and you. Just completely flexible in a way that was just completely new to me. Coming from, having worked with chart tools basically previously to, to visualize things.

[00:15:04] David: That makes a lot of sense. One last question, before we dig into the code not, so you talked about how, obviously the BBC has the BBC world service and their organ or, produced in different languages. Do you do anything now where are charts produced in different languages at all?

[00:15:20] In other words, for example, can you like make a chart in R in English and then iterate to make versions in different languages? Is that a thing at all?

[00:15:29] Nassos: Yeah. Yeah. Charts in Russian charts in a few other languages. We also had a kind of, yeah, a hurricane map. Making kind of script wasn't, didn't quite make it to package that we could also do into different languages. So it would take the data based on kind of which side of the world the hurricane was from and yeah, it needed a lot of dedication from world service, language colleagues whose. lives was made significantly easier by again, it's of that reversioning process that we are talking about, but on their end to get a graphic done from English into 30 languages. What, so we did two things to make that happen. One it was a lot of experimentation with kind of Unicode and character sets and all that.

[00:16:19] In our, that a lot of the, like went over my head, but was perfectly understood by people who have these issues with character sets every day. So a lot of world service language colleagues that were interested in data and graphics took a lot of that on. And then the second bit was trying to, for example, the Russian service or kind of other colleagues in the Americas hub, which is and visualizations for BBC Mundo and BBC, which the Spanish version and Brazilian as well.

[00:16:46] So also kind of part of the tools and the tutorials and everything that we developed. We also used it as an opportunity to that they joined. We put together a sort of course that was how they could get to a stage in R where they could make graphics. So it worked also. Yeah.

[00:17:08] Part of the work that we'd done was also to try and skill people up in R as well as that next step into using it for graphics. If that makes. Yeah.

[00:17:19] David: Yeah, definitely.

[00:17:20] I know you also made a sort of cookbook to help people learn about how to make different types of visualizations. I'm wondering if you can talk briefly about how that worked.

[00:17:30] Nassos: I think the cookbook was kind. As important, I think for us in terms of here's a repository of all this information and people could add their new chart styles to people could add tips and tricks to and it has been over the years added to again, internally as well in the and new joiners straight away have something. And we've had a few new people joining over the years. It's okay. So I think someone recently is very much a Python person, but because of the cookbook and because of the package could really quickly come into despite kind of being mainly Python literate.

[00:18:06] They could within week two of being at the BBC, given that they knew how to code and everything was there, they could of produce graphics in R and it just having the cookbook there to talk through how this all works and why. And what bit of code does what rather than having to sit and explain it with someone really helped. I.

[00:18:27] Clara: Yeah, I think that's totally right. And I think that it it was a really useful way of sharing information internally to feel like everybody was adding in their own. It was the idea was that it, and it did actually grow organically. The more people that kind of started using are, and started adding their own little kind of tips and tricks that they discovered along the way.

[00:18:45] And everybody was contributing to it.

[00:18:49] David: Yeah, cuz I know you've written that having both the theme as well as the cookbook really spurred additional interest in R and I think those are giving people that ownership and, letting people contribute. It seems like it would help with that as well as of course, like you said, giving people something to start with and some examples that they can use when they're getting started.

[00:19:09] Last question is if you have any advice for people, say someone is at an organization and they're thinking, oh, I should make a custom theme for my organization. Any advice on how to do that? Both, both from a code perspective, as well as from a organizational perspective.

[00:19:25] Clara: I guess the thing that I would have to say based on our experiences is to see the value of and not be afraid to experiment and just not to be afraid to to just sort. Of try things out and that it's okay if you don't have a huge plan from day one with everything set in stone and thinking, okay, so I'm gonna, I'm gonna do all these things.

[00:19:43] I'm gonna make this package. You can just start by just being curious, just having this one little problem that you have to solve and, go going. Okay. I figured that out. What would be the next step to improve? And then one day maybe, you go on from there, you solve the nice problem.

[00:19:56] One day you realize that you have custom theme set. Doesn't, it doesn't need to be more complicated than that.

[00:20:05] Nassos: Yeah I guess it's I guess the first question would be to anyone thinking about it is, do you have a need for it rather than start? And we, yeah, exactly as class said, so we didn't start with, let's make a package. It just built onto that. So it's kind. break things down into here's what I want to do.

[00:20:27] And here's what would benefit my organization and do that rather than start off with the solution is I want to write a package or I want to do a theme or, yeah, it's what do you actually want to do, want to produce? What sort of, what do you want your output to be? Cuz your output isn't unless you're in software development, your output ISN.

[00:20:49] I wanna do a package. So what is your output? And then find the building blocks that would provide a solution to that really. And yeah I guess from my experience is keep it super simple. I think what we ended up doing was having lots of people in a weird way, this helped.

[00:21:10] Other parts of our skillset development as well in that it was probably the first time we were U as a sort of team worked, 2018 sort of thing, using GitHub and in a way that everyone was pushing to the same repo and doing changes and commenting, and it really helped having.

[00:21:31] Loads of people working on one thing collaboratively two kind of really. It wasn't, it was never a task for one person go out and build a package. And then by little things that everyone added, we also worked better. Collaboratively, worked better. In terms of documentation worked better, we learned get hub, pull requests and all that stuff a little better.

[00:21:52] So it, it benefited a lot of things in parallel.

[00:21:55] David: Yeah. that makes a ton of sense. Great. Is there anything I didn't ask you about the theme or the process or anything like that you think would be useful?

[00:22:05] Nassos: Yeah. So a lot of the coronavirus look up pages that have that, that we've been running run off, are and produce graphics. I can send you a link to them, but it's this whole process of plugging in data, getting it, getting the data and using the package to produce.

[00:22:25] Graphics has lived on to a massive degree since then and continues like it's just, it's not something. We developed, we released and we let go. It's we've used it. I don't think there's been a day where someone at the BBC hasn't used the package to produce a graphic, I don't think.

[00:22:45] And definitely over the last sort of two years where we've been running coronavirus graphics every day really. And it was only due yeah, we can only do. in this way because of how kind of one step of it, which is building something as a custom made BBC theme using our is no longer something we think about.

[00:23:06] It's something that just happens cuz it's all there.

[00:23:10] David: Yeah.

[00:23:11] Nassos: so yeah. Yeah. I think that I don't think it's too big a statement to say that there's not been a day gone that someone hasn't produced. Yeah. I think that is probably true over the. Two three years. Definitely. So it has become a fundamental part of the BBC data journalism team's output, really.

[00:23:31] And the other thing is how many other people outside of the strictly data journalism team have taken to use it as well? I think that's a really yeah if, I guess if we'd be proud as a team of anything. That from my point of view is how it has basically, if you tell someone like, would you like to learn R everyone would say yes, but then how many people get lost along the journey, because.

[00:23:56] There is no, why am I doing this in their mind? And what can I get out of it? That is worth the pain, because it is a pain at first. But having showing someone like, or someone seeing like someone from another team, it's you know what, this graphic that a few kind of months ago you would have to do in this old process. If you. learn a few things or devote a bit of time a day. Here's the five lines of code that you can run and you can do that and you don't need help from, you can do it yourself. It's all contained within this. So I think that kind of spurred people a lot people outside kind of the data journalism team itself to take a real interest and kind of continue, keep on going because there is that, here's what I will produce at the end of it.

[00:24:47] That graphic and here, and you can see the code and here it's six lines of code. You can do that. Obviously they slowly learned that the tricky bit is the bits in between, which is like, how do I get my data to a stage to put into that graphic? And then it's okay, I've done the graphic.

[00:25:01] How do I add all these layers onto it that make it sing? But yeah, still it, yeah it really.

[00:25:07] David: Yeah. That's great. I was gonna ask where folks can learn more about you. Obviously folks can learn more about the, or see the package in action by going to the BBC website and seeing everything produced there. If they wanna learn more about you in your work, where would be the places to go to do that up?

[00:25:25] Nassos: Yeah, I guess the stuff that the data team does, we don't have a sort of page as a kind of data team page, but yeah, if any, anything to learn about the package or the cookbook, we did a medium post at the time, which talks about the process. So if there's anything that we might not have covered here that would be useful or a bit of a deep dive into the process and the code itself they should be in that medium post as well as links to the GitHub and the cookbook.

[00:25:50] And in terms of kind of work, yeah. It's it, yeah. Stuff that we do will show up across the BBC website. Why

[00:25:59] David: Yeah. And I know. So just if they wanna also find out more about you personally, Nassos I know your website is https://www.nassosstylianou.com/.

[00:26:11] Nassos: Yeah, sure. yeah, I, yeah. Yeah, Twitter, although I don't, yeah, I don't use it too much recently. Yeah. Yeah, Twitter and I've got linked to my website and the projects I've worked on and I've detail yeah, what works been done in R for each project as well. So that might be useful to see from the projects.

[00:26:28] What Yeah, especially, if someone who's interested in learning are there, there is a little bit at the bottom of each section, which says what I did and then what yeah. Language. Yeah.

[00:26:37] David: Okay. And then Clara looks like yours. Your website is

[00:26:41] Clara: Yeah, I do. I do have a website as well as I've not been very good at keeping it updated, but I'll need to look into that. But yeah, it is there. You can see some examples of I'm pretty sure you can. There are some links to, to with Charts that, that use beauty plot in there.

[00:26:57] David: Clara gbo.com. Both. It's yeah. Great.

[00:26:58] Clara: and you can find me on Twitter as well,

[00:27:01] David: Okay.

[00:27:01] Clara: Semi-active

[00:27:04] David: yeah, I'll include links to all those places. So folks can find out more about you, the work you've done and connect. Great. Thanks Clara. Thanks Nassos. I appreciate both of you spending some time with me today. And yeah. Thank you.

[00:27:17] Clara: yeah. Thank you.

[00:27:18] Thanks again for listening. I hope you found this conversation. Interesting. If you have any feedback, I'd love to hear it, David, at our, for the rest of us.com. Thanks.