st: RE: Splitting a data set by string labels stored in a variable

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

This is a -reshape- followed by some renaming 

. reshape wide value , i(regname year) j(varname) string
. renpfix value 

In Stata 12

. rename (value*) (*)

is an alternative to the -renpfix- statement. 

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

Love Hansson

I am currently doing some research on a dataset which contains
time-series data on a number of variables clustered by geographic
region. However, the dataset is structured in a way that stores the
the observations on all variables in a single value variable, and
keeps a separate name variable with the names of all the variables. To
illustrate, this is the format of the data:

varname    regname    year    value
Label1     Region1    2000    x1
Label1     Region1    2001    y1
Label1     Region2    2000    z1
Label1     Region2    2001    a1
Label2     Region1    2000    x2
Label2     Region1    2001    y2
Label2     Region2    2000    z2
Label2     Region2    2001    a2

Now, this is quite inconvenient, and I am therefore looking for an
easy way to convert this into a data file using the values of the
variable "varname" in the example above as variable names and mapping
the contents of "value" onto these, producing a result like the
following:

regname    year    Label1    Label2
Region1    2000    x1        x2
Region1    2001    y1        y2
Region2    2000    z1        z2
Region2    2001    a1        a2


*
*   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/