Inspired by #SharkWeek

Did you know there’s a shark themed badge on Trailhead? It’s a project about hunting prospects and closing deals lightning fast. And it has a shark fin circling money on the badge, so that’s rad. But I’m not a closer, I’m a builder, and I’ve been watching #sharkweek all week on Discovery. Have you?! I really loved their new show, Shark Academy, where 8 people were training to join a shark expedition. And I loved Brad Paisley playing his guitar under water – that was super fun.

One of my least favorite shows was the Jackass one. The whole time I was like WHAT ARE THEY DOING SOMEONE IS GOING TO GET HURT and then during the very last stunt, someone got bit by a shark. Now obviously this attack was provoked by a guy literally jumping over some sharks on water skis, but it got me wondering how other attacks happen and how often. So what did I do? I went looking for data! I found this data from Global Shark Attack File that’s pretty interesting, AND it already included the attack from the Jackass stunt!

Data Prep

After I downloaded the file it was pretty obvious pretty quickly that there was a lot of cleaning to do! Not gonna lie, I definitely didn’t do all of it, but if you’re interested in playing with it, here’s the list of stuff that I thought needed to be done:

• Clean and transform the report date – some of the rows had open text and none of it was actually formatted as a date.
• Clean and transform location data – it is listed as countries, but not in a standardized way. Countries needed to be matched and uniform and ultimately, updated to the way they are listed on my map components.
• Clean and transform activity data – this was actually pretty okay, but you could tell they were trying to standardize after a lot of years of just writing whatever.
• Clean up the sex column – it should have been just M and F, but some rows had extra spaces, some had ‘M x2’, etc.
• Evaluate which columns were filled in enough to use – a lot of columns, like ‘Time’ were not very complete. Very few rows had this data filled in, so I knew right away that I didn’t want to use it, even though that would have been cool to look at. Another column that would have been really cool was the ‘Species’ column! Unfortunately, there were a lot of blanks and what was there was not actually the name of a species, it was more like “2-3 foot white shark”.
• Injury could probably be cleaned up to body part(s) involved or generalized to just Fatal or not. Or split this into two columns for both.

The final columns I ended up using from the dataset were:

• Date
• Country
• Sex
• Activity
• Type
• Injury

To try not to be completely insensitive while I’m “playing” with this data, I also removed watercraft accidents and other large disasters.

Choosing a Color Palette

The first thing I did when I started building was find a rad shark gif because I knew I wanted to add some pizzazz to the dashboard! Once I picked my favorite one, I took a screenshot of it and uploaded it into the color palette generator on Canva so that I could have matching colors that looked great. If you haven’t used this tool before, I totally recommend it. I know at least for me, using the same old blues and brand colors gets boring and sometimes I just want to mix it up!

Layout

When I’m building a new dashboard, I usually like to put together some sort of wire frame before I actually start querying the data. This makes it faster to move things around and get a feel for what the final product may look like. I knew I wanted to really pop “above the fold”, or before you need to start scrolling down. I think I made that happen! Since my gif was sort of small, I set the dashboard width to 800 pixels so that my gif would always look great and not stretched out.

The Disclaimer and Then Some Data

Keep in mind here that I am not a scientist nor have I ever studied sharks. My entire shark education has come from watching Shark Week every year and so my perspective of this data is completely from a place of having fun and getting to use this awesome shark gif in a dashboard. I also have no idea how accurate or complete this dataset is that I downloaded, so please, don’t make any decisions about ocean activities or sharks by looking at this dashboard! Also keeping in mind that there are a lot of fatalities in this data, I hope I don’t come across as super insensitive and it’s definitely not my intention to offend anyone.

Again, not knowing anything about actual sharks, I just started playing with the data I had. Group by something, see if its interesting and then move on. I settled on a few different questions I wanted to try and answer:

1. Have there been more or less attacks over a certain period of time?
2. Are attacks typically provoked like in the case of the Jackass stunts, or not?
3. What activities were people doing when the attack happened?
4. What is the most common injury?
5. Which sex has the most attacks?
6. Where do attacks happen the most?

Keeping it simple and keeping it fun (and remembering it probably doesn’t really mean anything), here’s what I put together!

Maybe next time a #Sharknado dashboard? Or maybe I should stick to my business data 🙂