I wrote 114 posts for The DO Loop blog in 2015. Which were the most popular with readers?
In general, highly technical articles appeal to only a small group of readers, whereas less technical articles appeal to a larger audience. Consequently, many of my popular articles were related to data analysis or general SAS programming. These are topics that are relevant to all SAS programmers. However, it is gratifying to see that a few of my articles about computational statistics also attracted many readers.
General SAS techniques
Although I mostly write about statistical programming, several articles about how to do things in Base SAS were very popular:
- It was fun to write a point-counterpoint article in which three fictitious programmers argue about the merits of using mnemonic versus symbolic operators in SAS.
- What is the best way to suppress ODS output in SAS? I compare and contrast the NOPRINT option, the ODS EXCLUDE statement, and the ODS CLOSE statement.
- Do you ever submit a batch job that contains a long-running SAS program? It is often useful to break up the program into a sequence of shorter-running programs. You can use the -SET command line option and the &SYSGET macro to pass parameters to a SAS program.
- A lot of SAS users were happy to learn that there is an easy way to clear your SAS titles.
Statistical Graphics and Data Visualization
Everyone likes to learn better ways to visualize data in SAS! The following posts generated some interesting discussions.
- My most popular post in 2015 describes what happens when an uncapped bottle of nail polish slips out of a young girl's hand and stains the wall. The spiral stain and the subsequent statistical analysis of the splatter pattern must be seen to be believed!
- Visualizing the causes of airline crashes might sound macabre, but this article emphasizes that a simple data visualization can be more effective than a complex infographic.
- It is easy to plot a time series in PROC SGPLOT, but suppose you want to overlay a dozen line plots? It is useful to know how to transform data from wide to long form so that you can use a single SERIES statement to overlay the lines.
- A waterfall chart is a way to visualize the response of many patients in a clinical study. This popular article shows how to create a waterfall chart in SAS.
No list would be complete without including articles about how to implement statistical techniques, computations, simulations, and visualizations, which are the bread and butter of many SAS programmers. Here are a few topics that resonated with readers in 2015:
- Simulating data is an important skill for statistical programmers. SAS can be blindingly fast at simulating data if you follow the tips in this video presentation of ten tips for simulating data with SAS.
- Odds ratios are used frequently in clinical studies, but the standard visualization of odds ratios can be misleading. This article demonstrates that using a logarithmic scale for an odds ratio plot is often more understandable than the usual odds ratio plot.
- Regression models have assumptions. You can use PROC SGPLOT and other familiar procedures to visualize the assumed error distributions of a linear regression model.
- The Monty Hall Problem is a classic brain-teaser in conditional probability. Often people do not believe the result until they see a simulation. This article shows how you can simulate the Monty Hall Problem in SAS in less than 10 lines of code.
Did you make a New Year's resolution to improve your SAS skills this year? Start your new year by (re-)reading one of these 12 popular posts from 2015.
Do you remember a favorite article that didn't make the list? Name your favorite in the comments.