Hey there,

Thank you all so much for the positive feedback you all gave me after last week’s newsletter and yesterday’s YouTube video. I really can’t express what it means to me. Nor can I express how happy I am to be “back”. Because of my long break, your interests may have changed since you originally subscribed to the newsletter. Feel free to hit this button to subscribe so I’m not clogging up your inbox.

I strongly believe that you can and should be the person to analyze your own data. You shouldn’t need to hand your data off to someone else to analyze who likely doesn’t understand the collection of questions that interest you. But I can appreciate that this is a bit like a professional artist telling an aficionado that instead of stalking Sotheby’s they should be producing their own art. This reminds me of a quote from Ira Glass that I love:

You owe it to yourself to read the full quote linked above. This is a cool twist on my adage that you’re going to suck, but you’ll get better.

If you’ll hang with me on my art metaphor, I think there’s another important difference between an art aficionado and many of us. We may actually lack what Glass calls, “good taste”. We may not even know what is good and bad. I mean, I’m sure Jackson Pollock is great, but I just don’t get it.

What is “good taste” in data science? Last week, I introduced two characters. John who is an aspiring data scientist and Peggy who is John’s supervisor. Many of you who emailed me told me how you were a John, but now you’re a Peggy and you don’t have the ability to keep your skills current. Peggy may have a visceral reaction (good or bad!) to John’s code even if she doesn’t know what is good or bad about the code. John may be watching videos I put up or code that others in the lab have been writing to develop their own code. But it’s a bit like a Pollock painting - all over the place. They’re sure there’s something in there if we look at it long enough.

There’s actually a phrase used by computer programmers to describe this phenomenon - “code smells”. These are coding approaches that may work just fine, but … smell. They perhaps point to bigger problems or may cause something undesired to happen. You might be able to let it sit there for now, but it’ll likely be a problem soon. Think of a baby. You heard them just dropped a bomb in their diaper. But you are in the middle of making dinner. You know that they can sit in it for now, but soon everyone's going to be crying. If you want to see a great talk on code smells in R, check out Jenny Bryan’s talk the 2018 useR! Conference.

I contend that even if you don’t know how to program (Hi Peggy!), you can be trained to have good taste. If you’re John, you would also like to develop some good taste. I’ll be going through some of the code smells I look for in my code and those I collaborate with in the coming weeks.

Here’s one to get us started...

I frequently find library() function calls scattered throughout an R script. The script may run just fine. But, it can be confusing. If you run library(ggplot2) midway through your 500 line script (another smell!), but want to build a plot in the first part of the script, you’ll run into problems. Also, if you share your code with your colleague, it won’t be immediately clear to them what packages they need to have installed. The solution to this smell is to put your all of you library() calls at the top of the script. I encourage you to look through your code and see whether you can find this smell in your own work. Feel free to look around at other people’s code and see if you can spot this smell in the wild.

I would love it if you could reply to this email with any code smells that you notice in the code you look at. Are there any smells you are struggling to clean up?

Workshops

I'm pleased to be able to offer you one of three recent workshops! With each you'll get access to 18 hours of video content, Pat's code, and other materials. Click the buttons below to learn more

In case you missed it…

I posted my first YouTube video in 16 months yesterday. It’s the conclusion of the Drought Index visualization. I’m excited about the next project we’ll be starting on Monday!

Finally, if you would like to support the Riffomonas project financially, please consider becoming a patron through Patreon! There are multiple tiers and fun gifts for each. By no means do I expect people to become patrons, but if you need to be asked, there you go :)

I’ll talk to you more next week!

Pat