Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Command for selecting values


From   Nick Cox <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: Command for selecting values
Date   Wed, 30 Oct 2013 19:14:32 +0000

Your previous question was about cross-checking within companies.

But your earlier definitions imply that you calculated quintiles
separately for each value of -yrmonth- in which case your calculation
might be

bysort co_id yrmonth (quintile) : replace quintile = quintile[1]


Nick
[email protected]


On 30 October 2013 18:47, Clarice Martins <[email protected]> wrote:
> Dear Nick and Group,
>
> Reviewing the dataset, suggestion code and other material, I noticed that:
>
> - it is necessary to -replace- quintile data to other observations, respecting variable co_id (which the command did perfectly), but also taking into consideration the variable yrmonth.
>
> - i also figured out that my sample dataset was not very representative of the reality, since it was too simplistic (my apologies...!!!!!)
>
> - i created below new sample data, and tested now the command proposed created same output in main dataset and sample data ==> Results were the same in both
>
> *-------output from Stata----------------
> clear
> input byte co_id str4 ticker str10 rtype str10 yrmonth double return byte quintile
>
>         co_id     ticker       rtype     yrmonth      return  quintile
>   1. 1 ABCB formation jun2000 0.01 1
>   2. 1 ABCB formation jul2000 0.01 2
>   3. 1 ABCB formation aug2000 0.01 5
>   4. 1 ABCB buysell jul2000 0.01 .
>   5. 1 ABCB buysell aug2000 0.01 .
>   6. 1 ABCB buysell sep2000 0.01 .
>   7. 1 ABCB holding jan2001 0.01 .
>   8. 1 ABCB holding feb2001 0.01 .
>   9. 1 ABCB holding mar2001 0.01 .
>  10. 2 AEDU formation jun2000 0.01 2
>  11. 2 AEDU formation jul2000 0.01 1
>  12. 2 AEDU formation aug2000 0.01 3
>  13. 2 AEDU buysell jul2000 0.01 .
>  14. 2 AEDU buysell aug2000 0.01 .
>  15. 2 AEDU buysell sep2000 0.01 .
>  16. 2 AEDU holding jan2001 0.01 .
>  17. 2 AEDU holding feb2001 0.01 .
>  18. 2 AEDU holding mar2001 0.01 .
>  19. 3 AMBV formation jun2000 0.01 3
>  20. 3 AMBV formation jul2000 0.01 4
>  21. 3 AMBV formation aug2000 0.01 5
>  22. 3 AMBV buysell jul2000 0.01 .
>  23. 3 AMBV buysell aug2000 0.01 .
>  24. 3 AMBV buysell sep2000 0.01 .
>  25. 3 AMBV holding jan2001 0.01 .
>  26. 3 AMBV holding feb2001 0.01 .
>  27. 3 AMBV holding mar2001 0.01 .
>  28. 4 FFFF formation jun2000 0.01 4
>  29. 4 FFFF formation jul2000 0.01 3
>  30. 4 FFFF formation aug2000 0.01 5
>  31. 4 FFFF buysell jul2000 0.01 .
>  32. 4 FFFF buysell aug2000 0.01 .
>  33. 4 FFFF buysell sep2000 0.01 .
>  34. 4 FFFF holding jan2001 0.01 .
>  35. 4 FFFF holding feb2001 0.01 .
>  36. 4 FFFF holding mar2001 0.01 .
>  37. 5 BBBB formation jun2000 0.01 4
>  38. 5 BBBB formation jul2000 0.01 2
>  39. 5 BBBB formation aug2000 0.01 1
>  40. 5 BBBB buysell jul2000 0.01 .
>  41. 5 BBBB buysell aug2000 0.01 .
>  42. 5 BBBB buysell sep2000 0.01 .
>  43. 5 BBBB holding jan2001 0.01 .
>  44. 5 BBBB holding feb2001 0.01 .
>  45. 5 BBBB holding mar2001 0.01 .
>  46. 6 MMMM formation jun2000 0.01 .
>  47. 6 MMMM formation jul2000 0.01 .
>  48. 6 MMMM formation aug2000 0.01 .
>  49. 6 MMMM buysell jul2000 0.01 .
>  50. 6 MMMM buysell aug2000 0.01 .
>  51. 6 MMMM buysell sep2000 0.01 .
>  52. 6 MMMM holding jan2001 0.01 .
>  53. 6 MMMM holding feb2001 0.01 .
>  54. 6 MMMM holding mar2001 0.01 .
>  55. end
>
>
>
> . tabulate quintile, missing
>
>    quintile |      Freq.     Percent        Cum.
> ------------+-----------------------------------
>           1 |          3        5.56        5.56
>           2 |          3        5.56       11.11
>           3 |          3        5.56       16.67
>           4 |          3        5.56       22.22
>           5 |          3        5.56       27.78
>           . |         39       72.22      100.00
> ------------+-----------------------------------
>       Total |         54      100.00
>
>
>
> . bysort co_id (quintile) : replace quintile = quintile[1]
> (40 real changes made)
>
> .
>
>
> . tabulate quintile, missing
>
>    quintile |      Freq.     Percent        Cum.
> ------------+-----------------------------------
>           1 |         27       50.00       50.00
>           3 |         18       33.33       83.33
>           . |          9       16.67      100.00
> ------------+-----------------------------------
>       Total |         54      100.00
>
> .
>
> - i do need to keep the 2, 4, and 5 as well,
>
>
> Thanks again!!!!  Learning a lot... in this group!!!
>
> Clarice
>
>
>
> On Oct 30, 2013, at 11:32 AM, Nick Cox wrote:
>
>> . bysort co_id (quintile) : replace quintile = quintile[1]
>>
>> lets you "spread" the non-missing values of -quintile- to all
>> observations for each company. You can then select appropriate
>> companies freely.
>>
>> For an introductory review of several tricks in this territory, see
>>
>> SJ-11-2 dm0055  . . . . . . . . . . . . . .  Speaking Stata: Compared with ...
>>        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
>>        Q2/11   SJ 11(2):305--314                                (no commands)
>>        reviews techniques for relating values to values in other
>>        observations
>>
>> Nick
>> [email protected]
>>
>>
>> On 30 October 2013 13:04, Clarice Martins <[email protected]> wrote:
>>
>>> I am using Stata 12.1 for Mac.
>>>
>>> My problem is:
>>> - within a dataset (with the structure below), I need to -if rtype=="formation" & yrmonth == "`lev'" & quintile==1- , pick values from -rtype==buysell- but from same companies that are in the -if- selection.
>>>
>>> I thought of something like this for a code:
>>>
>>> *-------begin code------
>>> gen P1_sell = .
>>> levelsof yrmonth, local(levs)
>>>
>>>
>>> foreach lev of local levs {
>>>        egen OK = anymatch(co_id) values (if rtype=="formation" & yrmonth == "`lev'" & quintile==1) //na aceita if ai dentro
>>>        replace P1_sell = return if rtype=="buysell" & yrmonth == "`lev'" & co_id = anymatch(OK)
>>>        drop temp
>>> }
>>> *-------end code------------
>>>
>>> But I know there is something missing to tell the code to pick values from other observations. I have been researching similar cases on the net, but no luck yet.
>>>
>>> Can anyone give me a hint???
>>>
>>> Thank you very much,
>>> Clarice
>>>
>>> *----dataset sample----
>>> clear
>>> input byte co_id str4 ticker str10 rtype str10 yrmonth double return byte quintile
>>> 1 ABCB formation jun2000 0.01 1
>>> 1 ABCB formation jul2000 0.01 1
>>> 1 ABCB formation aug2000 0.01 1
>>> 1 ABCB buysell jul2000 0.01 .
>>> 1 ABCB buysell aug2000 0.01 .
>>> 1 ABCB buysell sep2000 0.01 .
>>> 1 ABCB holding jan2001 0.01 .
>>> 1 ABCB holding feb2001 0.01 .
>>> 1 ABCB holding mar2001 0.01 .
>>> 2 AEDU formation jun2000 0.01 2
>>> 2 AEDU formation jul2000 0.01 2
>>> 2 AEDU formation aug2000 0.01 2
>>> 2 AEDU buysell jul2000 0.01 .
>>> 2 AEDU buysell aug2000 0.01 .
>>> 2 AEDU buysell sep2000 0.01 .
>>> 2 AEDU holding jan2001 0.01 .
>>> 2 AEDU holding feb2001 0.01 .
>>> 2 AEDU holding mar2001 0.01 .
>>> 3 AMBV formation jun2000 0.01 3
>>> 3 AMBV formation jul2000 0.01 3
>>> 3 AMBV formation aug2000 0.01 3
>>> 3 AMBV buysell jul2000 0.01 .
>>> 3 AMBV buysell aug2000 0.01 .
>>> 3 AMBV buysell sep2000 0.01 .
>>> 3 AMBV holding jan2001 0.01 .
>>> 3 AMBV holding feb2001 0.01 .
>>> 3 AMBV holding mar2001 0.01 .
>>> 4 FFFF formation jun2000 0.01 4
>>> 4 FFFF formation jul2000 0.01 4
>>> 4 FFFF formation aug2000 0.01 4
>>> 4 FFFF buysell jul2000 0.01 .
>>> 4 FFFF buysell aug2000 0.01 .
>>> 4 FFFF buysell sep2000 0.01 .
>>> 4 FFFF holding jan2001 0.01 .
>>> 4 FFFF holding feb2001 0.01 .
>>> 4 FFFF holding mar2001 0.01 .
>>> 5 BBBB formation jun2000 0.01 5
>>> 5 BBBB formation jul2000 0.01 5
>>> 5 BBBB formation aug2000 0.01 5
>>> 5 BBBB buysell jul2000 0.01 .
>>> 5 BBBB buysell aug2000 0.01 .
>>> 5 BBBB buysell sep2000 0.01 .
>>> 5 BBBB holding jan2001 0.01 .
>>> 5 BBBB holding feb2001 0.01 .
>>> 5 BBBB holding mar2001 0.01 .
>>> 6 MMMM formation jun2000 0.01 .
>>> 6 MMMM formation jul2000 0.01 .
>>> 6 MMMM formation aug2000 0.01 .
>>> 6 MMMM buysell jul2000 0.01 .
>>> 6 MMMM buysell aug2000 0.01 .
>>> 6 MMMM buysell sep2000 0.01 .
>>> 6 MMMM holding jan2001 0.01 .
>>> 6 MMMM holding feb2001 0.01 .
>>> 6 MMMM holding mar2001 0.01 .
>>> end
>>> *-------end--------
>>>
>> *
>> *   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index