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 at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: Creating a dummy variable under certain conditions


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Creating a dummy variable under certain conditions
Date   Sun, 3 Jun 2012 20:06:45 +0100

It makes no sense to work with L.gvkey as this just means the previous
panel identifier, which is constant within a panel . This has already
been pointed out earlier in the thread. Referring to L.gvkey was a
slip that was quickly corrected.

Nick

On Sun, Jun 3, 2012 at 7:59 PM, sebas nicaise <sebasnicaise@hotmail.com> wrote:
> Still not working :(
>
>
>
> When I use this :
>
> bys gvkey gen new_viol_dummy=L.gvkey+L2.gvkey+L3.gvkey+L4.gvkey==0
>
>
> bys gvkey:replace new_viol_dummy=0 if new_viol_dummy[_n-1]==1
>
>
> I get this:
>
> date gvkey viol new_viol_dummy
> 1997q1 1004 0      0
> 1997q2 1004 0      0
> 1997q3 1004 0      0
> 1997q4 1004 0      0
> 1998q1 1004 0      1
> 1998q2 1004 0      0
> 1998q3 1004 0      1
> 1998q4 1004 0      0
> 1999q1 1004 0      1
> 1999q2 1004 0      0
> 1999q3 1004 0      1
> 1999q4 1004 0      0
> 2000q1 1004 0      1
>
>
>
>
>
>
>
> Could you have another look at it :)?
>
>
>
>
>
>
>
>
>
>
>
>
>
>> Date: Sun, 3 Jun 2012 20:24:33 +0200
>> Subject: Re: st: Creating a dummy variable under certain conditions
>> From: abhimanyu.arora1987@gmail.com
>> To: statalist@hsphsun2.harvard.edu
>>
>> Now I see, you have an additional condition, Sebastian.
>>
>> In that case, you could simply add an additional command in addition
>> to either Nick's or my solution with L. (not the F. one)
>> -bys gvkey:replace new_viol_dummy=0 if new_viol_dummy[_n-1]==1
>>
>>
>> date gvkey viol new_viol_dummy
>> 1997q1 1004 0 0
>> 1997q2 1004 0 0
>> 1997q3 1004 0 0
>> 1997q4 1004 0 0
>> 1998q1 1004 0 0
>> 1998q2 1004 0 0
>> 1997q1 1021 0 0
>> 1997q2 1021 0 0
>> 1997q3 1021 0 0
>> 1997q4 1021 0 0
>> 1998q1 1021 0 0
>> 1998q2 1021 0 0
>> 1998q3 1021 0 0
>> 1998q4 1021 1 1
>> 1999q1 1021 1 0
>> 1999q2 1021 1 0
>> 1999q3 1021 1 0
>> 1999q4 1021 0 0
>>
>>
>>
>>
>> On Sun, Jun 3, 2012 at 7:46 PM, sebas nicaise <sebasnicaise@hotmail.com> wrote:
>>> Dear Abhimanyu,
>>>
>>> This also does not give me the right answer.It gives:
>>>
>>> date gvkey viol new_viol_dummy1998q4 1095 0 01999q1 1095 0 01999q2 1095 0 01999q3 1095 0 01999q4 1095 1 02000q1 1095 0 02000q2 1095 1 12000q3 1095 1 12000q4 1095 1 12001q1 1095 1 12001q2 1095 1 12001q3 1095 1 0
>>>
>>> It should be:000010000000
>>>
>>> I think the command should be more complicated since the new viol dummy variable is only 1 under certain strict conditions.NEWVIOL is only 1 if the VIOL dummy of the corresponding GVKEY has a 1 IF in the four previous quarters a 0. Furthermore, if NEWVIOL documents a 1, it can only be a 1 again after four quarters of 0.
>>>
>>> Hope you, or someone else, can help me with this seemingly easy task :)
>>>
>>> Thank you,Sebastiaan
>>> ----------------------------------------
>>>> Date: Sat, 2 Jun 2012 17:21:54 +0200
>>>> Subject: Re: st: Creating a dummy variable under certain conditions
>>>> From: abhimanyu.arora1987@gmail.com
>>>> To: statalist@hsphsun2.harvard.edu
>>>>
>>>> Okay, if I infer correctly, try this: bys gvkey: gen
>>>> new_viol_dummy=viol+F1.viol+F2.viol+F3.viol==4.
>>>> Seems good?
>>>>
>>>>
>>>>
>>>> On Sat, Jun 2, 2012 at 5:07 PM, sebas nicaise <sebasnicaise@hotmail.com> wrote:
>>>>> Maybe I can better show it visually:
>>>>> Your commands give me the following output:
>>>>> bys gvkey: gen new_viol_dummy=L.viol+L2.viol+L3.viol+L4.viol==0
>>>>> date gvkey viol new_viol_dummy 1997q1 1004 0 01997q2 1004 0 01997q3 1004 0 01997q4 1004 0 01998q1 1004 0 11998q2 1004 0 11997q1 1021 0 01997q2 1021 0 01997q3 1021 0 01997q4 1021 0 01998q1 1021 0 11998q2 1021 0 11998q3 1021 0 11998q4 1021 1 11999q1 1021 1 01999q2 1021 1 01999q3 1021 1 01999q4 1021 0 0
>>>>> While I want this: date gvkey viol new_viol_dummy 1997q1 1004 0 01997q2 1004 0 01997q3 1004 0 01997q4 1004 0 01998q1 1004 0 01998q2 1004 0 01997q1 1021 0 01997q2 1021 0 01997q3 1021 0 01997q4 1021 0 01998q1 1021 0 01998q2 1021 0 01998q3 1021 0 01998q4 1021 1 11999q1 1021 1 01999q2 1021 1 01999q3 1021 1 01999q4 1021 0 0
>>>>> If it is not clear what I mean, let me know :)
>>>>> Thank you,Sebastiaan
>>>>>> Date: Sat, 2 Jun 2012 16:34:41 +0200
>>>>>> Subject: Re: st: Creating a dummy variable under certain conditions
>>>>>> From: abhimanyu.arora1987@gmail.com
>>>>>> To: statalist@hsphsun2.harvard.edu
>>>>>>
>>>>>> Sebas,
>>>>>> Adding to Nick's comment, did you not -xtset- the data? You also have
>>>>>> an option of bys id_var: command, otherwise
>>>>>>
>>>>>>
>>>>>> On Sat, Jun 2, 2012 at 4:29 PM, Nick Cox <njcoxstata@gmail.com> wrote:
>>>>>>> You will probably want to exclude or ignore the first few observations
>>>>>>> in each panel or -replace- it with missing. Otherwise, what's wrong?
>>>>>>>
>>>>>>> Abhimanyu's slip has already been corrected in this thread, as below.
>>>>>>>
>>>>>>> Nick
>>>>>>>
>>>>>>> On Sat, Jun 2, 2012 at 3:12 PM, sebas nicaise <sebasnicaise@hotmail.com> wrote:
>>>>>>>
>>>>>>>> Thank you for the quick replys.
>>>>>>>>
>>>>>>>> However, the suggested solutions did not give the right answer.
>>>>>>>>
>>>>>>>> Nick, your solution gives me way too many 1's. I do think the first step with the cusum was the right way to go.
>>>>>>>>
>>>>>>>> Abhimanyu, your solution does not distinquish between the id's, the calculation has to start over for every gvkey.
>>>>>>>>
>>>>>>>>
>>>>>>>> Hope you can have another look at it, I will also try to work with the suggestions you gave me.
>>>>>>>>
>>>>>>>>> From: abhimanyu.arora1987@gmail.com
>>>>>>>>> To: statalist@hsphsun2.harvard.edu
>>>>>>>>>
>>>>>>>>> Thanks Nick, yes.
>>>>>>>>> But I would go for your solution which would be better in case he had
>>>>>>>>> a longer lag.
>>>>>>>>>
>>>>>>>>> On Sat, Jun 2, 2012 at 2:56 PM, Nick Cox <njcoxstata@gmail.com> wrote:
>>>>>>>>>> This looks OK as long as you substitute -viol- for -gvkey-. The
>>>>>>>>>> previous values concerned are for -viol-, not -gvkey-.
>>>>>>>>>>
>>>>>>>>>> Nick
>>>>>>>>>>
>>>>>>>>>> On Sat, Jun 2, 2012 at 1:51 PM, Abhimanyu Arora
>>>>>>>>>> <abhimanyu.arora1987@gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> Try this if your data is -xtset-ted
>>>>>>>>>>> gen new_viol_dummy=L.gvkey+L2.gvkey+L3.gvkey+L4.gvkey==0
>>>>>>>>>>
>>>>>>>>>>> On Sat, Jun 2, 2012 at 2:06 PM, sebas nicaise <sebasnicaise@hotmail.com>
>>>>>>>>>>>
>>>>>>>>>>>> I am struggeling with creating a new dummy variable. I have a very large
>>>>>>>>>>>> (unbalanced) panel dataset that looks like this:
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> date gvkey viol
>>>>>>>>>>>> 1997q1 1021 0
>>>>>>>>>>>> 1997q2 1021 0
>>>>>>>>>>>> 1997q3 1021 0
>>>>>>>>>>>> 1997q4 1021 0
>>>>>>>>>>>> 1998q1 1021 1
>>>>>>>>>>>> 1998q2 1021 0
>>>>>>>>>>>> 1998q3 1021 0
>>>>>>>>>>>> 1998q4 1021 1
>>>>>>>>>>>> 1999q1 1021 0
>>>>>>>>>>>>
>>>>>>>>>>>> 1999q2 1021 0
>>>>>>>>>>>>
>>>>>>>>>>>> 1999q3 1021 0
>>>>>>>>>>>> 1999q2 1023 0
>>>>>>>>>>>> 1999q3 1023 1
>>>>>>>>>>>> 1999q4 1023 0
>>>>>>>>>>>> 2000q1 1023 0
>>>>>>>>>>>> 2000q2 1023 1
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Where date is calender quarter, gvkey is the company identifier and viol a
>>>>>>>>>>>> dummy which equals 1 if the firm is in violation of a covenant in a private
>>>>>>>>>>>> debt contract.
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> The problem is that I have to create a new dummy variable ( new violation)
>>>>>>>>>>>> which meets certain conditions.
>>>>>>>>>>>>
>>>>>>>>>>>> The new violation dummy is only allowed to be 1 if in the previous 4
>>>>>>>>>>>> quarters of a certain company (gvkey) is 0.
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> This means for example that:
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> date gvkey viol Newviolation
>>>>>>>>>>>> 1997q1 1021 0 0
>>>>>>>>>>>> 1997q2 1021 0 0
>>>>>>>>>>>> 1997q3 1021 0 0
>>>>>>>>>>>> 1997q4 1021 0 0
>>>>>>>>>>>> 1998q1 1021 1 1
>>>>>>>>>>>> 1998q2 1021 0 0
>>>>>>>>>>>> 1998q3 1021 0 0
>>>>>>>>>>>> 1998q4 1021 1 0
>>>>>>>>>>>> 1999q1 1021 0 0
>>>>>>>>>>>>
>>>>>>>>>>>> 1999q2 1021 0 0
>>>>>>>>>>>>
>>>>>>>>>>>> 1999q3 1021 0 0
>>>>>>>>>>>> 1999q2 1023 0 0
>>>>>>>>>>>> 1999q3 1023 1 0
>>>>>>>>>>>> 1999q4 1023 0 0
>>>>>>>>>>>> 2000q1 1023 0 0
>>>>>>>>>>>> 2000q2 1023 1 0
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> I do not know how to tell Stata to create this variable, Hope someone is
>>>>>>>>>>>> so kind to help me.
>>>>>>>>>>
>>>>>>>>>> *
>>>>>>>>>> * For searches and help try:
>>>>>>>>>> * http://www.stata.com/help.cgi?search
>>>>>>>>>> * http://www.stata.com/support/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/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/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/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/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/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/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/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/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/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/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index