Re: st: complete data import

[Douglas Levy <douglas_levy@post.harvard.edu>]

Sergiy,
Thanks for the suggestion, Unfortunately, these are raw cash register
sales so there is no file further upstream.
-Doug

On Mon, Apr 7, 2014 at 8:44 PM, Sergiy Radyakin <serjradyakin@gmail.com> wrote:
> where record types are not numeric but strings "ITM" "TND" and "MEM",
> but the fact that it is a CSV means, perhaps, that someone has already
> played with this file, since in all the CSPro files I've seen so far
> the storage was positional. Perhaps you can find the raw file
> somewhere upstream, then employ
> -usecspro- to read it in:
> http://ideas.repec.org/p/boc/usug13/17.html
>
> (slide #9 illustrates the dataset layout of CSPro that is remarkably similar).
>
> Hope this helps.
>
> Best, Sergiy Radyakin
>
> On Mon, Apr 7, 2014 at 3:45 PM, Douglas Levy
> <douglas_levy@post.harvard.edu> wrote:
>> Hi All,
>> I have a complicated .csv file that I'm trying to import and I'm
>> hoping you can provide some guidance on the best strategy. See the
>> example below. Each record in the file consists of up to 3 line types
>> labeled "MEM", "ITM", and "TND". Each line type has a particular
>> format. For example, "MEM" line types have var1, var2, and var3. "ITM"
>> has var4, var5, and var6. Etc. Each record may or may not have a "MEM"
>> line,  has 1 or more "ITM" lines, and has 1 "TND" line that ends the
>> record. A record is defined as ending with the "TND" line. The "MEM"
>> line may come at any point in the record, but is often first.
>>
>> Example data
>> ITM,S,0,1,1.35,5100001261,soup,
>> TND,N,0,0,0,1.35,0,0
>> ITM,S,0,1,1.35,5100001261,soup,
>> TND,N,0,0,0,1.35,0,0
>> MEM,N,280211400001,,,,,
>> ITM,S,0,1,1.35,5100001261,soup,
>> TND,N,0,0,0,0,11.28,0
>> ITM,S,0,1,1.29,5000001011,milk,
>> TND,N,0,0,0,1.29,0,0
>> MEM,N,280211400159,,,,,
>> ITM,S,0,1,1.29,5000001011,milk,
>> ITM,S,0,1,1.29,5000001011,milk,
>> ITM,S,0,1,1.29,5000001011,milk,
>> TND,N,4,0,0,0,0,0
>>
>> I'd like to attach a unique record identifier the "TND" line and I'd
>> like to attach the "TND" and "MEM" info for a given record to each
>> "ITM" line in the record. Ultimately, I'd like the Stata file for the
>> above data to look like this, where the number after "soup" or "milk"
>> is the record identifier (placement of that variable is not
>> important):
>> ITM,S,0,1,1.35,5100001261,soup, 1, TND,N,0,0,0,1.35,0,0
>> ITM,S,0,1,1.35,5100001261,soup, 2, TND,N,0,0,0,1.35,0,0
>> ITM,S,0,1,1.35,5100001261,soup, 3, TND,N,0,0,0,0,11.28,0,
>> MEM,N,280211400001,,,,,
>> ITM,S,0,1,1.29,5000001011,milk, 4, TND,N,0,0,0,1.29,0,0
>> ITM,S,0,1,1.29,5000001011,milk, 5, TND,N,4,0,0,0,0,0, MEM,N,280211400159,,,,,
>> ITM,S,0,1,1.29,5000001011,milk, 5, TND,N,4,0,0,0,0,0, MEM,N,280211400159,,,,,
>> ITM,S,0,1,1.29,5000001011,milk, 5, TND,N,4,0,0,0,0,0, MEM,N,280211400159,,,,,
>>
>> Any advice would be most welcome!
>> Best,
>> Doug
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/faqs/resources/statalist-faq/
>> *   http://www.ats.ucla.edu/stat/stata/
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/