Re: st: Creating a contact tracing file (levelsof & nested foreach loops)

[Nick Cox <njcoxstata@gmail.com>]

Have a look at

SJ-7-3  pr0033  . . . . . . . . . . . . . .  Stata tip 51: Events in intervals
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q3/07   SJ 7(3):440--443                                 (no commands)
        tip for counting or summarizing irregularly spaced
        events in intervals

which is accessible to all at

http://www.stata-journal.com/sjpdf.html?articlenum=pr0033

On Fri, Sep 28, 2012 at 4:40 PM, Puddicombe, David
<David.Puddicombe@bccdc.ca> wrote:

> I would like to create a file that indicates which cases of an infectious disease outbreak may have been the source of infection for later cases.
>
> The variables in my dataset are
> id_n (non-identifying id, values 1 to 83)
> date_onset (%td date of onset of disease variable)
> city (city where each case resides, numeric variable)
>
> Cases occurred in seven cities:
>
>   City      |      Freq.     Percent        Cum.
> ------------+-----------------------------------
>      city_1 |         46       55.42       55.42
>      city_2 |          2        2.41       57.83
>      city_3 |         12       14.46       72.29
>      city_4 |          3        3.61       75.90
>      city_5 |         18       21.69       97.59
>      city_7 |          1        1.20       98.80
>      city_8 |          1        1.20      100.00
> ------------+-----------------------------------
>       Total |         83      100.00
>
> Within each city, I would like to create indicator variables (0 No 1 Yes) to describe which of the subsequent cases may have been infected by each case and to generate variables with the id number of each potential subsequent case.
>
> The rules to define whether a case might be the source of infection for a subsequent case are:
> they are in the same city; and
> 7<=(date_onset[case2] - date_onset[case1])<=18
>
> For example, case 1 in city_1 could be the source of infection for the other 45 cases in city_1.  I would like to create two variables for each of these 45 potential contacts(potential_contact_n_yn; potential_contact_n_id). For case 2 in city_1, there are 44 potential subsequent cases etc.
>
> I've tried several levelsof and foreach loops but cannot make my code work.  My aim is to create a wide dataset describing the relationships between cases within each city and then transpose the file from wide to long with one row for each potential source of infection to use in a network analysis program (cytoscape).
>
> Any suggestions/help would be greatly appreciated.  Some sample data are below.
>
> id_n date_onset city
> 1    09mar2010  city_1
> 2    10mar2010  city_1
> 3    11mar2010  city_1
> 4    18mar2010  city_1
> 5    18mar2010  city_1
> 6    18mar2010  city_1
> 7    18mar2010  city_1
> 8    19mar2010  city_1
> 9    19mar2010  city_1
> 10   20mar2010  city_1
> 11   21mar2010  city_1
> 12   21mar2010  city_1
> 13   22mar2010  city_1
> 14   22mar2010  city_1
> 15   22mar2010  city_1
> 16   22mar2010  city_1
> 17   22mar2010  city_2
> 18   24mar2010  city_1
> 19   25mar2010  city_3
> 20   25mar2010  city_7
> 21   25mar2010  city_1
> 22   25mar2010  city_1
> 23   26mar2010  city_1
> 24   26mar2010  city_1
> 25   26mar2010  city_3
> 26   27mar2010  city_3
> 27   28mar2010  city_5
> 28   28mar2010  city_3
> 29   29mar2010  city_1
> 30   29mar2010  city_1
>

*
*   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/