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

# Re: st: Differencing problem

 From Nick Cox <[email protected]> To "[email protected]" <[email protected]> Subject Re: st: Differencing problem Date Thu, 27 Jun 2013 15:49:23 +0100

```As they say in London, but for different reasons,

Mind the gap!

As far as d. is concerned, you have no value for 1999, so it can only
shrug its shoulders and return missing.

Set delta to 10 years, and everyone will be happy.
Nick
[email protected]

On 27 June 2013 15:31, Michael Betz <[email protected]> wrote:
> Sure. Here it is:
>
> . xtset cbsa year
>        panel variable:  cbsa (strongly balanced)
>         time variable:  year, 1990 to 2000, but with gaps
>                 delta:  1 unit
>
> The only two years in the data are 1990 and 2000.
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of Jeph Herrin
> Sent: Thursday, June 27, 2013 8:59 AM
> To: [email protected]
> Subject: Re: st: Differencing problem
>
> Can you share the output of -xtset- ?
>
>
> On 6/27/2013 8:09 AM, Michael Betz wrote:
>> Right. Whenever I've done this before that is exactly what I would get too. Now the value of the difference for the second observation is missing in each instance. Not sure what is going on.
>>
>> -----Original Message-----
>> From: [email protected]
>> [mailto:[email protected]] On Behalf Of Nick Cox
>> Sent: Thursday, June 27, 2013 4:26 AM
>> To: [email protected]
>> Subject: Re: st: Differencing problem
>>
>> Nothing obviously wrong there.
>>
>> With panels length 2 what I expect is what I get, just one value of the difference for the second observation.
>>
>> . set obs 10
>> obs was 0, now 10
>>
>> . gen y = runiform()
>>
>> . egen id = seq(), block(2)
>>
>> . egen t = seq(), to(2)
>>
>> . xtset id t
>>         panel variable:  id (strongly balanced)
>>          time variable:  t, 1 to 2
>>                  delta:  1 unit
>>
>> . gen dy = d.y
>> (5 missing values generated)
>>
>> . l
>>
>>       +-------------------------------+
>>       |        y   id   t          dy |
>>       |-------------------------------|
>>    1. | .1369841    1   1           . |
>>    2. | .6432207    1   2    .5062366 |
>>    3. | .5578017    2   1           . |
>>    4. | .6047949    2   2    .0469932 |
>>    5. |  .684176    3   1           . |
>>       |-------------------------------|
>>    6. | .1086679    3   2    -.575508 |
>>    7. | .6184582    4   1           . |
>>    8. | .0610638    4   2   -.5573944 |
>>    9. | .5552388    5   1           . |
>>   10. | .8714491    5   2    .3162102 |
>>       +-------------------------------+ Nick [email protected]
>>
>>
>> On 27 June 2013 02:56, Michael Betz <[email protected]> wrote:
>>> . d college_
>>>
>>>                storage  display     value
>>> variable name   type   format      label      variable label
>>> ------------------------------------------------------------------------------------------------------------------------------------
>>> college_        float  %9.0g
>>>
>>> . su college_
>>>
>>>      Variable |       Obs        Mean    Std. Dev.       Min        Max
>>> -------------+-------------------------------------------------------
>>> -------------+-
>>>      college_ |       716 94970.41    257294.9       3479    3707827
>>>
>>> -----Original Message-----
>>> From: [email protected]
>>> [mailto:[email protected]] On Behalf Of Nick Cox
>>> Sent: Wednesday, June 26, 2013 6:53 PM
>>> To: [email protected]
>>> Subject: Re: st: Differencing problem
>>>
>>> I would bet No. Show us the results of
>>>
>>> d var
>>> su var
>>>
>>> Nick
>>> [email protected]
>>>
>>> On 26 June 2013 23:38, Michael Betz <[email protected]> wrote:
>>>
>>>> The code you sent works fine. I am using -xtset- to define the panel data. Will that make a difference?
>>>
>>>
>>>> See if this works. The problem may be with the time variable. Sergiy
>>>>
>>>> sysuse auto,clear
>>>> generate t=_n
>>>> tsset t
>>>> generate x=d.mpg
>>>> list mpg x
>>>
>>> On Wed, Jun 26, 2013 at 4:20 PM, Michael Betz <[email protected]> wrote:
>>>
>>>>> I am trying to use the "d." operator to get differenced
>>>>> observations but am having some trouble. When I try to calculate
>>>>> differenced observations using
>>>>>
>>>>> gen newvar=d.(var)
>>>>>
>>>>> Stata returns each observation as missing. My data has observations for each variable in each of two time periods and variation exists over the two time periods. I've used the "d." before without problems so I'm not sure what I'm missing here. Any help is appreciated.
>>>
>>> *
>>> *   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/
>>
>>
>>
>> *
>> *   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/

*
*   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/
```