|
|
|
|
| |
|
|
| Cox, Nicholas J.. |
| Table-like graphs can be interesting, useful, and even mildly innovative. This column outlines some Stata techniques for producing such graphs. graph dot is likely to be the most under-appreciated command among all existing commands. Using by() with various choices is a good way to mimic a categorical axis in many graph commands. When graph bar or graph dot is not flexible enough to do what you want, moving to the more flexible twoway is usually advisable. labmask and seqvar are introduced as new commands useful for preparing axis labels and axis positions for categorical variables. Applications of these ideas to, e.g., confidence interval plots lies ahead. |
| Tipo: Article |
Palavras-chave: Labmask; Seqvar; Tables; Graphs; Dot charts; Research Methods/ Statistical Methods. |
| Ano: 2008 |
URL: http://purl.umn.edu/122591 |
| |
|
|
| Cox, Nicholas J.. |
| Three user-written commands are reviewed as illustrations of different approaches to tabulation problems, each one step beyond what is possible to do directly through official Stata. tabcount is a wrapper for tabdisp written to produce tables that show how often specified values occur or specified conditions are satisfied so that, in particular, tables may include explicit zeros whenever desired. makematrix is designed for situations in which a table of results may be compiled by populating a matrix. matrix list or list may then be used to display the table. groups shows frequencies of combinations of values using list. Users should find these commands to be helpful additions to the toolkit. Programmers may be interested in examples of the wrapper... |
| Tipo: Journal Article |
Palavras-chave: Tables; Matrices; Tabcount; Makematrix; Groups; Tabdisp; List; Research Methods/ Statistical Methods. |
| Ano: 2003 |
URL: http://purl.umn.edu/116197 |
| |
|
|
|
