Re: st: changing values of a string variable

[Nick Cox <njcoxstata@gmail.com>]

II agree with what I take your main point. As I said, I don't use this
syntax, but it seems to be what Tashi was expecting or reaching out
for.

Nick

On Tue, Apr 24, 2012 at 9:56 AM, Prakash Singh <prakashbhu@gmail.com> wrote:
> Thanks Nick for putting my words in a neat way.
>
> I am not convince why use the command
> replace firm1 = . if firm1 != "id_131":firm1
> when you can use
> replace firm1 = . if firm1 != 1 (suppose the value correnponding to
> the value lebel id_131 is 1 which one can easly get to know)
>
> Prakash
>
> On Tue, Apr 24, 2012 at 1:47 PM, Ronnie Babigumira <rb.glists@gmail.com> wrote:
>> Nick, this is neat, thanks.
>>
>> Tashi, this is certainly closer to what you had in mind
>>
>> encode firms, gen(firm1)
>> replace firm1 = . if firm1 != "id_131":firm1
>>
>> Ronnie
>>
>> --
>> 010100100110111101101110011011100110100101100101
>>
>>
>>
>>
>> On Tuesday, April 24, 2012 at 9:52 AM, Nick Cox wrote:
>>
>>> Prakash's opening statement is incorrect. If you -encode- a string
>>> variable, it loses no property and indeed stays exactly as it is. What
>>> happens is that a numeric variable is created with value labels equal
>>> to the string values.
>>>
>>> Tashi appears to be seeking something like the last line here
>>>
>>> sysuse auto
>>> label def rep78 1 appalling 2 adequate 3 average 4 amazing 5 astounding
>>> label val rep78
>>>
>>> . count if rep78=="average":rep78
>>> 30
>>>
>>> For more, see in [U] and e.g.
>>>
>>> SJ-4-4 dm0009 . . . . . . . Stata tip 14: Using value labels in expressions
>>> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . K. Higbee
>>> Q4/04 SJ 4(4):488--489 (no commands)
>>> tips for using value labels in expressions
>>>
>>> Ken Higbee's paper is accessible at
>>>
>>> http://www.stata-journal.com/sjpdf.html?articlenum=dm0009
>>>
>>> This syntax has been in Stata a very long time. I don't use it, partly
>>> because I keep forgetting about it and partly because I would need to
>>> look it up to remember what it is, and my wild guess is that many
>>> experienced users would say something similar.
>>>
>>> Nick
>>>
>>>
>>> Prakash Singh [edited slightly]
>>>
>>> > when you -encode- any variable it loses its string property so
>>> > there would be numeric value corresponding to id_131 in firms1 and you
>>> > should use that value in your command line
>>>
>>>
>>>
>>> tashi lama
>>>
>>> > > I found matters similar to what I am going to post have been been discussed in the past but I am not sure if I understood. Here is my dataset
>>> > >
>>> > > hits firms
>>> > > 3 id_6429
>>> > > 10 id_131
>>> > > 8 id_2837
>>> > >
>>> > > 12 id_90
>>>
>>> > > The problem is to have sth like this
>>> > >
>>> > > hits firms
>>> > >
>>> > > 3 .
>>> > >
>>> > > 10 id_131
>>> > > 8 .
>>> > >
>>> > > 12 .
>>> > >
>>> > >
>>> > > so I said 1. encode firm, gen(firm1)
>>> > >
>>> > > replace firm1=. if firm1!=id_131
>>> > >
>>> > > drop firms
>>> > >
>>> > > I also tried 2. replace firms="" if firms!="id_131"
>>>
>>> > > It didn't give me the expected output although I feel like my codes are correct at least in principle. Any idea?
*
*   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/