st: re: destring

[Marcello Pagano <pagano@hsph.harvard.edu>]

Marcello Pagano wrote:

> For Kit:
>
> Suzy said
>
> Since the majority of the datapoints are actually numeric rather than
> alpha-numeric...is there a way for me to modify the 11 variable (dxcodes
> and procodes) from string to numeric and then modify the alpha-numeric
> datapoints within that variable to be read as either a string or numeric
> (I don't care which as long as it can be read)... and as the other
> gentlemen had stated - it's probably best to retain the code as is
> rather than drop the dashes and alphabets.
>
>
> Per my previous message, if you look closely at what encode is doing, 
> you will see that it does what you need. encode makes a string 
> variable into a numeric variable with string value labels. Thus the 
> variable can be displayed by its original contents, just as the 
> variable foreign in auto.dta appears to have string content--but it 
> also can be used in routines which require a numeric representation 
> (such as anova or regress). So read the variables that have both pure 
> numeric and alphanumeric values as strings--that is the only way that 
> all of their contents can possibly be retained within Stata--and 
> encode them. That's all it takes. They cannot be 'modified from string 
> to numeric' without destroying the alphanumeric content.
>
> Kit
>
>
> *
> *   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/
*
*   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/