We find this to be the easiest approach.Īnother approach is to export the results to Excel.Įxport excel 'c:\tables\means.xlsx', firstrow(variables) Since every observation has the same values for the three means, we can keep just one observation.Īt this point we can browse the data, copy, and paste it to an empty Excel file template we've created with column and row labels, colors, and all the other formatting we want. Next put them into a matrix, and then create a variable for each cell in the matrix: We can compare these values with the means printed in the table above and see that we have the estimates of the means. After that we'll talk about a number of user-contributed commands that work with results from estimation commands.įirst, let's verify that we have in fact selected the right matrix of estimates: We'll demonstrate one way that uses official Stata commands and is flexible, but it requires you to manage the data a bit more than you might care to do. We want to create variables that we can export to Excel, so we first need to put the means into variables that won't go away when we run another estimation command. The estimates of the means are contained in a matrix called e(b), which has dimensions 1 x 3. Here's a subset of the estimate parameters stored temporarily after the svy:mean command that produced the table above:įor purposes of this example, we're only interested in the means. You can look at what results have been stored using either return list or for estimation commands ereturn list.
![stata 13 command list stata 13 command list](https://image1.slideserve.com/1591098/stata-commands-in-recoding-variables-l.jpg)
These temporary variables will continue to store these values until you use another Stata command that replaces those results, or until you end your Stata session. Most Stata commands produce temporary variables containing the key results. We can do this be taking advantage of a very nice feature of Stata. Suppose that we want to make a table of just the means in the above example.