I'll try the easier part of this.
Once your filename is in
pl[1]
you can refer to that on the fly by
`=pl[1]'
so your -merge- command would then be
merge obsno using `=pl[1]', keep(pl)
The command
replace pl ="10u5uc" if _merge==1
would more generally be
replace pl = pl[1] if _merge == 1
if I understand you correctly.
Nick
[email protected]
Juan Solon
> I would like to advise on using a data dictionary to extract comments
> from a .txt file and putting this into a variable.
>
> I wanted to read data from a series of text files, with each file
> containing unique file identifiers/descriptors in a header in lines
> 1-10. All files are named data.txt and I would have to do this
> repeatedly and thus, Im looking for a solution that replicable across
> files. The data itself is in a tabular format and starts on
> line 11 and
> would look like this:
> Spot counts:
> 1 2 3 4 5 6 7 8 9 10 11 12
> A - - - - - - - - - - - -
> B - - - - - - - - - - - -
> C - - - - - - - - - - - -
> D - - - - - - - - - - - -
> E - - - - - - - - - - 197 9
> F - - - - - - - - - - 188 7
> G - - - - - - 3 2 1 204 189 78
> H - - - - - - 1 2 0 254 195 63
>
> I have made a do file that can read the tabular data
> beginning in Line
> 11 contained in data.txt using infile and a data dictionary.
> My problem
> is that I want to tag each observation with a unique identifier that
> would be taken from line 1. This is important because I would
> eventually merge all the data from the text files and I would need to
> know the source.
>
> The solution I tried seems very clumsy, and it doesn't
> actually work as
> I wanted it to. I created a separate file with a variable (pl)
> containing the unique file identifier (the string from Line
> 1). I then
> generated an observation no using generate obsno=_n and saved the file
> with the filename as the string from line 1. This file can then be
> merged with the tabular data .
>
>
> - start of do file -
> set more off
> infile using eli2.dct, clear
> drop in 2/l
> gen pl=substr(plate,18,6) /* this is data from the first line for
> example a string "10u5uc"*/
> gen obsno=_n
> local file =pl
> sort obsno
> save `file', replace
> infile using eli.dct, clear
> gen obsno=_n
> sort obsno
> merge obsno using 10u5uc, keep(pl) /* How do I do this using a macro
> instead of typing "10u5uc"
>
> - end of do file
>
>
> The resulting merged file then contains 8 observations and a string
> variable pl. As expected, variable pl is missing for records 2-8. I
> would like variable pl for records 2-8 to have the same value
> as record
> 1. I can do this manually:
>
> replace pl ="10u5uc" if _merge==1
>
> but would rather that it were a macro. At this point, I have
> hit a brick
> wall!!!
>
> 1. How do I extract the data in Line 1 into a variable and
> repeat this
> for all the records ?
> 2. Is it possible to merge files and refer to the using
> dataset using a
> macro?
*
* 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/
