When you have data whose range spans several orders of magnitude, you should consider whether a log transform will enhance the visualization. The standard visualization technique to use in this situation is the logarithmic transformation of data. To visualize those observations (while not losing information about Chris and Michelle) requires some sort of transformation that distributes the data more uniformly within the plot. But what about the other nameless markers near the origin? You can see that there are many people who have posted between 10 and 30 comments, but the current plot makes it difficult to find out who they are. provided that your goal is to highlight the people who post the most comments and classify them into "commenter" or "responder." The plot enables you to identify about a dozen of the 50 people in the data set. The scatter plot is an excellent visualization of the data. The big outlier is Chris, who has initiated almost 100 original comments while also posting more than 500 responses to comments on his popular blog. In contrast, Sanjay and Robert are regular SAS bloggers and most of their remarks are responses to other people's comments. For example, Michelle and Tricia (lower right) often comment on blogs, but few of their comments are in response to others. The line, which was drawn by using the LINEPARM statement, enables you to see who has initiated many comments and who has posted many responses. Those who have commented more than 30 times are labeled, and a line is drawn with unit slope. The scatter plot shows the number of comments and responses for 50 people. Scatter x=Comment y=Response / datalabel=TruncName The following call to the SGPLOT procedure create a scatter plot of these data: I consider only commenters who have posted more than ten comments. You canÄownload the program that creates the data and the graphs in this article. Glad you liked it!" is classified as a response. Thanks!" is classified as a comment, whereas "You're welcome. For each comment, he recorded the name of the commenter and whether the comment was an original comment or a response to a previous comment. My colleague, Chris Hemedinger, wrote a SAS program that collects data about comments on the Web site. It also discusses a common problem: How to transform data that range over several orders of magnitudes but that also contain zeros? (Recall that the logarithm of zero is undefined!) This article shows several ways to create a scatter plot with logarithmic axes in SAS and discusses some of the advantages and disadvantages of using logarithmic transformations. If you are trying to visualize numerical data that range over several magnitudes, conventional wisdom says that a log transformation of the data can often result in a better visualization.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |