2018 programming goals: SAS, R, python

Featured image: “Midnight,” digital art I made in 2014. Moiré exclusion patterns are created from layering multiple photographs of a LCD screen. Bands of solid color are from certain non-proportionally stretched layers in the vertical direction.

Hi all! It’s been a hot minute since I updated here. A lot of big events have happened, ranging from good to terrible, since the last post…

I got engaged, my fiancé and I got mugged and assaulted (yup! that happened), and my rabbit has been sick. We are ok now, planning a courthouse wedding, and my rabbit is doing much better. Recently, inadvertently reformatted my HTPC with 2 TB data on it, but it’s all for fun. I advise against upgrades from Ubuntu 16 to 17, and against having a roof destroyed that may cause rain to fall on said machine. One of the drives is still iffy.

In my free moments I’ve been studying SAS harder in order to take the Base Certification exam soon. I work part-time for my mom, who runs a S corp business for SAS programming. I actually find SAS is quite enjoyable, and the learning curve was much less severe than R. From what I can tell, SAS is still going strong in the pharmaceutical industry due its dedicated support and warranty. R, while free, is without a doubt harder to pick up on the fly and comes with “ABSOLUTELY NO WARRANTY” as stated in the loading message. For larger companies and/or labs, SAS may be preferred over R in order to reduce individuals’ time coding from scratch, to make code deployment more reliable. Really, in general, the less programmers need write new code to complete day-to-day tasks, the more efficient the whole team is. Makes sense. So, I am working my way through the official SAS e-learning classes and taking on extra work tasks.

I also need to learn python. I chose Perl in graduate school, but python would have been a better choice I suspect. The community for python applied to data science seems much larger than for Perl, which, as of writing, I still see used for large, active software packages (i.e. ANNOVAR), but has scant application to data visualization. I think I will have no trouble picking up python 2 or 3 — I started on 3 this year but it was pushed to the margins. One attractive feature python has over R is more robust graphics engines than R ggplot. I’ve hit the “limit” of ggplot complex graphs numerous times in the past year, which was frustrating. It is as if R does not have the tools to make some imagined graphs a programmatic reality.

As for R, is will still be my daily-flyer language and I will continue to be asked to use it for pipeline development. My colleagues and coworkers at Johns Hopkins & elsewhere are most familiar with R.

How am I going to keep SAS and R separate, plus add in python? So far it has been doable. Even though the syntax and conceptual approaches to handling data are vastly different, I think for now I have enough RAM in my brain to keep them separate. I hope in 2018 it will continue to be this way. When I attempted to learn German in college, after ~8 years of French, I started interjecting French words into my German speaking! I suppose it will be easy for a debugger or interpreter to spot, in any case. Cheers, all!

Photo below: a couple years ago after receiving Razer peripherals (no longer have bangs)

Photo du 17-12-2014 à 6.10 PM #2

Comments are closed.

Powered by WordPress.com.

Up ↑

%d bloggers like this: