Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: How to drop missing values by period?


From   Maarten Buis <maartenlbuis@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: How to drop missing values by period?
Date   Mon, 11 Feb 2013 17:25:43 +0100

Did you actually try what Nick proposed? If you did than you would
have seen that what you said is not true. If you drop the observation
(row) "id 1, period 1" than the observation (row) "id 1, period 2"
will still be there.

*------------------ begin example ------------------
drop _all
input id      period  x1      x2
1       1       1       4
2       1       2       .
3       1       .       2
1       2       4       4
2       2       5       3
3       2       .       2
end

drop if missing(x1, x2)

list
*------------------- end example -------------------
(For more on examples I sent to the Statalist see:
http://www.maartenbuis.nl/example_faq )

Hope this helps,
Maarten

On Mon, Feb 11, 2013 at 5:16 PM, Xixi Lin <winnielxx@gmail.com> wrote:
> Hi Nick,
>
> I know what do you mean, if I, for instance, drop id 1, then in period
> 2, id 1 is also gone. I am thinking is that possible only to drop id 1
> in period 1, while in other periods, id 1 is still there. I am not
> sure did I clearify this time?
>
> Best,
> Xixi Lin
>
> On Mon, Feb 11, 2013 at 10:25 AM, Nick Cox <njcoxstata@gmail.com> wrote:
>> Thanks for the example. Stata usually takes care of this for you in that
>>
>> 1. A command uses as many as observations as possible.
>>
>> 2. Most statistical commands ignore missing values.
>>
>> But in your example
>>
>> . drop if missing(x1, x2)
>>
>> will drop observations if either -x1- or -x2- is missing. It's not
>> clear to me that you need to think of this period by period, as that
>> makes no difference to the outcome.
>>
>> Nick
>>
>> On Mon, Feb 11, 2013 at 3:18 PM, Xixi Lin <winnielxx@gmail.com> wrote:
>>> Hi Nick,
>>>
>>> Thanks for the instant reply. What I mean is like:
>>> id      period  x1      x2
>>> 1       1       1       4
>>> 2       1       2       .
>>> 3       1       .       2
>>> 1       2       4       4
>>> 2       2       5       3
>>> 3       2       .       2
>>>
>>> In period 1, I want to keep id 1 because id 2 and id 3 have missing
>>> values; in period 2, I want to keep id 1 &2, because id 3 has missing
>>> values.
>>>
>>> I have time series data, but I wanna do a cross-sectional regressions
>>> and wanna keep the maximum of observations. So I am wondering if it is
>>> possible to drop obs only in one period while keeping it in  another
>>> period.
>>>
>>> Thanks.
>>> Xixi Lin
>>>
>>> On Mon, Feb 11, 2013 at 9:59 AM, Nick Cox <njcoxstata@gmail.com> wrote:
>>>> This sounds to me as if you want -keep- and -drop- the same
>>>> observations. That can't be done.
>>>>
>>>> Perhaps you should show what your data look like with variable names,
>>>> and link your question to such an example.
>>>>
>>>> Nick
>>>>
>>>> On Mon, Feb 11, 2013 at 2:53 PM, Xixi Lin <winnielxx@gmail.com> wrote:
>>>>
>>>>> I tried to drop missing values period by period, in 120 periods. In other
>>>>> words, in period 1, I want to drop the observations if dependent variables
>>>>> have missing values only in period 1; while in period 2, drop the
>>>>> observations if dependent variables have missing values only in period 2,
>>>>> should keep those that missing in period 1 but not missing in period 2. For
>>>>> instance:
>>>>>
>>>>> In period 1, I want to keep id 1; in period 2, I want to keep id 1 &2.
>>>>>
>>>>> Does anyone knows how to do that? Thanks a lot.
>> *
>> *   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/



-- 
---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------
*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index