st: I'm sorry to do this again, but...

[Jon Ebeling <jebeling@rocko.lab.csuchico.edu>]

I have tried using a suggestion from some of the listees:


. generate hour2 = real (hours)
real not found
r(111);


Then I tried using :


. destring hours, replace
hours already numeric; no replace

While the first procedure seemed reasonable but would not yield what I 
want, the second is true, but it again does not yield what I want. I 
want to be able to see the numeric values when I run a tab on hours.

Here is what the underlying numeric values are:


. sort secnum

. by secnum: sum hours

----------------------------------------------------------------------------------------------------------
-> secnum = 1 cathy

   Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
      hours |         1          89           .         89         89

----------------------------------------------------------------------------------------------------------
-> secnum = 2 hassan

   Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
      hours |         1          12           .         12         12

----------------------------------------------------------------------------------------------------------
-> secnum = 3 freddi

   Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
      hours |         1          24           .         24         24

----------------------------------------------------------------------------------------------------------
-> secnum = 9

   Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
      hours |         1          23           .         23         23

I thought I could use a replace command for each but if the data base is 
large then I have a problem. Further, when I tried to use a replace 
command such as


. replace hours = 89 if secnum == 1
(0 real changes made)

the problem is in how to convert these foolish "na" responses to value 
labels. I'm afraid I can't figure it out. I do recall several years ago 
a procedure to do this in the labeling functions, but I can't recall how 
to do it.
At the present the tab of this variable produces the following:


. tab hours

 enter the |
 number of |
     hours |
written on |
       the |
questionnai |
 re in the |
space here. |
    enter  |      Freq.     Percent        Cum.
------------+-----------------------------------
        na |          1       25.00       25.00
        na |          1       25.00       50.00
        na |          1       25.00       75.00
        na |          1       25.00      100.00
------------+-----------------------------------
     Total |          4      100.00

What I want to do is drop the na and get the actual numbers behind them. 
The numeric values are there but they do not show on the labeling system.  
Here is how I got to this stage:

I use a software package called Ci3, and it has prcedures for locating 
andn producing numeric values. In this case I used num etc etc to locate 
the measure when entered.

then I converted the output to spss version 11 save file. Then I used 
stat transfer to convert it from spss to a stata 8 file. The result is this.

If anyone might know how to solve the problem I would appreciate it.

jon ebeling
csu, chico






*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/