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: hourly panel data
From
Nick Cox <[email protected]>
To
[email protected]
Subject
Re: st: hourly panel data
Date
Wed, 4 May 2011 01:31:22 +0100
Let's focus on using hours as a unit.
The range in hours for each id is naturally based on the difference
between the first and last times. Range here is the number of possible
observations.
bysort id (year day hour) : gen double range =
(24 * (mdy(1,1,year[_N]) + day[_N] - 1) + hour[_N])
-
(24 * (mdy(1,1,year[1]) + day[1] - 1) + hour[1])
+ 1
The number of observed times is easier
by id : gen observed = _N
and the number of gaps is just the difference.
A gap is not present in the data. Missings are missing values present
in the data.
You should be able to build on that. But it seems that you have many
days with no observations and it would seem futile to pad out the data
with 24 empty observations in each instance.
You don't explain what "monitor" means; perhaps that is what -id- identifies.
Nick
On Tue, May 3, 2011 at 10:31 PM, Amber Pearson <[email protected]> wrote:
> I am trying to tell stata that my data are a panel time series. I have the following variables: id (1-120), year (2005-2007), day (1-365), hour (11-15), and value (0-67.5)
>
> The datset looks like this:
>
> id year day hour value
> 1 2005 2 11 23.4
> 1 2005 25 12 14.5
> 1 2006 300 13 6.2
> 2 2005 151 11 44.0
> 2 2007 6 15 13.6
> 3 2006 45 12 11.3
>
>
> I want to be able to tell how many hours are missing for each day by monitor. I also want to be able to tell how many days are missing by monitor.
*
* 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/