st: RE: format (decimals) in results stored

[Nick Cox <n.j.cox@durham.ac.uk>]

Here is a dopey example.  

. sysuse auto
(1978 Automobile Data)

. su mpg

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
         mpg |        74     21.2973    5.785503         12         41

. histogram mpg , subtitle(mean `: di %2.1f r(mean)' mpg)
(bin=8, start=12, width=3.625)

Know that the text "mean 21.3 mpg" appears as subtitle. 

The incantation here is 

`: di %2.1f r(mean)' 

which is documented more generally at -help extended_fcn- as -display display_directive-. 

Let's replay that as a two-step. 

1. Given that you have r-class results in memory, you could go 

. di %2.1f r(mean)

to see the mean formatted as you wish. 

2. And you can also do that on the fly. Wrapping that in `:  ' means that Stata will pretend to do that in mid-air, but the command concerned will see the result. 

In practice, I often do this outside the main call. I do something like 

local text : di %2.1f r(mean) 

and invoke the macro within the graph command. 

In practice, decide how many decimal places you want, then add one. So use formats %2.1f, %3.2f, etc. Don't think: oh, those numbers will be about 1000, so I need %5.1f. That way, you will just get awkward spaces when the number is in fact smaller. 

Use something like %2.1f and Stata will just assign enough space automagically, unless the number is very greatly bigger. If your numbers are highly predictable, use something other than an f-format. 

Nick 
n.j.cox@durham.ac.uk 

Leandro Brufman

I am trying to include results stored in r() in graphs. For instance,
to add them as labels in graphs - with suboptions xlabel(0(1)10
`r(mean)' "Mean") -
The only problem I have is that r(mean) includes a lot of decimals. I
would like to change the format, but only to the `r(mean)' (ie, its
not helpful to change the dispaly format of the whole x - i want
integers from 0 to 10 and then the mean with two decimals)

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/