In a previous blog post, I showed how to use the LOGISTIC procedure to construct a receiver operator characteristic (ROC) curve in SAS. That same day, Charlie H. blogged about how to use the DATA step to construct an ROC curve from basic principles. It has been a long time
"Dad? How many times do I have to roll a die until all six sides appear?" I stopped what I was doing to consider my son's question. Although I could figure out the answer mathematically, sometimes experiments are more powerful than math equations for showing how probability works. "Why don't
One of the joys of statistics is that you can often use different methods to estimate the same quantity. Last week I described how to compute a parametric density estimate for univariate data, and use the parameters estimates to compute the area under the probability density function (PDF). This article
If you create a scatter plot of highly correlated data, you will see little more than a thin cloud of points. Small-scale relationships in the data might be masked by the correlation. For example, Luke Miller recently posted a scatter plot that compares the body temperature of snails when they
In a previous article, I discussed random jittering as a technique to reduce overplotting in scatter plots. The example used data that are rounded to the nearest unit, although the idea applies equally well to ordinal data in general. The act of jittering (adding random noise to data) is a
Jittering. To a statistician, it is more than what happens when you drink too much coffee. Jittering is the act of adding random noise to data in order to prevent overplotting in statistical graphs. Overplotting can occur when a continuous measurement is rounded to some convenient unit. This has the
The area under a density estimate curve gives information about the probability that an event occurs. The simplest density estimate is a histogram, and last week I described a few ways to compute empirical estimates of probabilities from histograms and from the data themselves, including how to construct the empirical
A reader commented to me that he wants to use the HISTOGRAM statement of the SGPLOT procedure to overlay two histograms on a single plot. He could do it, but unfortunately SAS was choosing a large bin width for one of the variables and a small bin width for the
Each Sunday, my local paper has a starburst image on the front page that proclaims "Up to $169 in Coupons!" (The value changes from week to week.) One day I looked at the image and thought, "Does the paper hire someone to count the coupons? Is this claim a good
In a previous blog post, I presented a short SAS/IML function module that implements the trapezoidal rule. The trapezoidal rule is a numerical integration scheme that gives the integral of a piecewise linear function that passes through a given set of points. This article demonstrates an application of using the
Many people know that the SGPLOT procedure in SAS 9.2 can create a large number of interesting graphs. Some people also know how to create a panel of graphs (all of the same type) by using the SGPANEL procedure. But did you know that you can also create a panel
A fundamental operation in data analysis is finding data that satisfy some criterion. How many people are older than 85? What are the phone numbers of the voters who are registered Democrats? These questions are examples of locating data with certain properties or characteristics. The SAS DATA step has a
For years I've been making presentations about SAS/IML software at conferences. Since 2008, I've always mentioned to SAS customers that they can call R from within SAS/IML software. (This feature was introduced in SAS/IML Studio 3.2 and was added to the IML procedure in SAS/IML 9.22.) I also included a
When Charlie H. posted an interesting article titled "Top 10 most powerful functions for PROC SQL," there was one item on his list that was unfamiliar: the COALESCE function. (Edit: Charlie's blog no longer exists. The article used to be available at http://www.sasanalysis.com/2011/01/top-10-most-powerful-functions-for-proc.html) Ever since I posted my first response,
Last week the Flowing Data blog published an excellent visualization of the flight patterns of major US airlines. On Friday, I sent the link to Robert Allison, my partner in the 2009 ASA Data Expo, which explored airline data. Robert had written a SAS program for the Expo that plots
More than a month ago I wrote a first article in response to an interesting article by Charlie H. titled Top 10 most powerful functions for PROC SQL. In that article I described SAS/IML equivalents to the MONOTONIC, COUNT, N, FREQ, and NMISS Functions in PROC SQL. In this article,
In last week's article on how to create a funnel plot in SAS, I wrote the following comment: I have not adjusted the control limits for multiple comparisons. I am doing nine comparisons of individual means to the overall mean, but the limits are based on the assumption that I'm
The log transformation is one of the most useful transformations in data analysis. It is used as a transformation to normality and as a variance stabilizing transformation. A log transformation is often used as part of exploratory data analysis in order to visualize (and later model) data that ranges over
Last week I showed how to create a funnel plot in SAS. A funnel plot enables you to compare the mean values (or rates, or proportions) of many groups to some other value. The group means are often compared to the overall mean, but they could also be compared to
In a previous blog post, I showed how you can use simulation to construct confidence intervals for ranks. This idea (from a paper by E. Marshall and D. Spiegelhalter), enables you to display a graph that compares the performance of several institutions, where "institutions" can mean schools, companies, airlines, or
At the beginning of 2011, I heard about the Dow Piano, which was created by CNNMoney.com. The Dow Piano visualizes the performance of the Dow Jones industrial average in 2010 with a line plot, but also adds an auditory component. As Bård Edlund, Art Director at CNNMoney.com, said, The daily
In a previous blog post about computing confidence intervals for rankings, I inadvertently used the VAR function in SAS/IML 9.22, without providing equivalent functionality for those readers who are running an earlier version of SAS/IML software. (Thanks to Eric for pointing this out.) If you are using a version of
When comparing scores from different subjects, it is often useful to rank the subjects. A rank is the order of a subject when the associated score is listed in ascending order. I've written a few articles about the importance of including confidence intervals when you display rankings, but I haven't
This week, I posted the 100th article to The DO Loop. To celebrate, I'm going to analyze the content of my first 100 articles. In December 2010, I compiled a list of The DO Loop's most-read posts, so I won't repeat that exercise. Instead, I thought it would be interesting
In a previous post, I described how to compute means and standard errors for data that I want to rank. The example data (which are available for download) are mean daily delays for 20 US airlines in 2007. The previous post carried out steps 1 and 2 of the method
I recently posted an article about representing uncertainty in rankings on the blog of the ASA Section for Statistical Programmers and Analysts (SSPA). The posting discusses the importance of including confidence intervals or other indicators of uncertainty when you display rankings. Today's article complements the SSPA post by showing how
Loony. Zany. Brilliant. Hysterical. Those are some of the adjectives I use to describe The Far Side® cartoons by Gary Larson from the 1980s and early '90s. I recently rediscovered an old book, The Far Side Gallery 2, which collects some of the best of Larson's wonderfully wacky cartoons. Every
Several times a year, I am contacted by a SAS account manager who tells me that a customer has asked whether it is possible to convert a MATLAB program to the SAS/IML language. Often the customer has an existing MATLAB program and wants to include the computation as part of
Do you have many points in your scatter plots that overlap each other? If so, your graph exhibits overplotting. Overplotting occurs when many points have similar coordinates. For example, the following scatter plot (which is produced by using the ODS statistical graphics procedure, SGPLOT) displays 12,000 points, many of which
I was inspired by Chris Hemedinger's blog posts about his daughter's science fair project. Explaining statistics to a pre-teenager can be a humbling experience. My 11-year-old son likes science. He recently set about trying to measure which of three projectile launchers is the most accurate. I think he wanted to
