Stata Table - Restrict Observations

[Hans Schubert]

Hello

Is is possible when working with TABLE in Stata to restrict observations to between two values of a variable....

Example/// entered into the window is the expression outofsampledata == 1 & won == 1 & myfinalprice >20 <30

In the above I want to restrict observations to between 20 and 30 of the 'myfinalprice' variable.
Have tried a few things but none are providing a solution.

Thanks.

[Clyde Schechter]

Not directly; -table- does not support -if- or -in- clauses. But what you can do is restrict the data set before calling -table-.

preserve
keep if outofsampledata == 1 & won == 1 & inrange(myfinalprice, 20, 30)
table whatever
restore

If you don't actually need the data that are dropped before -table- for anything else later, then you can skip the -preserve- and -restore- commands.

[Hemanshu Kumar]

I don't follow. Are we talking about the new -table- command in Stata 17? If so, the help clearly states:

Full syntax

table (rowspec) (colspec) [(tabspec)] [if] [in] [weight] [, options]

Edit: I just confirmed, and the old -table- command (pre-Stata 17) also used to allow for -in- and -if-.

[Rich Goldstein]

I find Clyde Schechter response confusing as -table- does support in and if; here is an example using the lbw data set

. table race low if inrange(age,20,25)

--------------------------------
        |    birthweight<2500g  
        |     N      Y     Total
--------+-----------------------
race    |                       
  white |    29     10        39
  black |     6      6        12
  other |    18     15        33
  Total |    53     31        84
--------------------------------


. table race low

--------------------------------
        |    birthweight<2500g  
        |      N      Y    Total
--------+-----------------------
race    |                       
  white |     73     23       96
  black |     15     11       26
  other |     42     25       67
  Total |    130     59      189
--------------------------------

the second table is show as evidence that the "if" actually did work

[Clyde Schechter]

Sorry, Rich is correct, and an -if- clause is all that is needed here. Apologies for the confusion.

[Hans Schubert]

Thanks all. Much appreciated.