Re: st: Drop under several conditions and time (year)

["Brill, Robert" <robertbrill@austin.utexas.edu>]

Your -if- subsetting seems to be incorrect. Right now, your code says
³drop if the year is equal to 2007, or, if the year is 2008 and long-term
debt is greater than total assets.²

What you intended to type was probably something more like:

drop if (Year==2007 & LongTermDebt>TotalAssets) | (Year==2008 &
LongTermDebt>TotalAssets)

But that will only remove the rows for which those statements are true,
not the entire firm.

Stata does not know the format of your data, so it will not drop the
entire firm, only the rows for which the -if- statement is true.

For this reason, previous answers to your question have used the -egen-
command to create a drop indicator, which will be one for all rows of a
firm if the firm fulfills a certain condition.

An easier way for you to understand your problem might be to reshape your
data into ³wide² format, such that each row represents one firm, and has a
variable for each variable for each year. Thus 13 rows of LongTermDebt are
converted to thirteen columns of LongTermDebt named LongTermDebt2000,
LongTermDebt2001, etc. You can do this by doing:

reshape wide LongTermDebt TotalAssets, i(Firm) j(Year)

drop if LongTermDebt2007>TotalAssets2007 |
LongTermDebt2008>LongTermDebt2008

reshape long

Of course, if your data contain variables other than those you have
described, you¹ll need to include them in the argument of reshape (before
the comma). The advantage of reshaping your data to wide is that you will
be able to drop observations on single row conditions, rather than trying
to understand -egen-, which can be complicated for a novice to data
management. 

Regarding your other question, if you simply have a list of keys you want
to drop, rather than encoding your string variable, you could simply use
the correct syntax for strings:

drop if Key==³C0056"

Notice the quotation marks‹values of string variables must be enclosed in
quotation marks for Stata to understand what to evaluate.








On 3/23/14, 4:10 PM, "Julian Kochan" <julian.kochan@gmx.de> wrote:

>Hi,
>
>i have a set of data with financial firm characteristics of about 10.000
>firms for Year 2000 - Year 2012.
>
>Let's assume: The first column of the data table displays the Name of the
>firm, the second the Year, the third displays the LongTerm Debt and the
>fourth the TotalAssets.
>Thus, 13 rows in a row (2000-2012) display information about just one
>firm - just the year varies.
>
>Problem: I want to drop those firms, for which LongTermDebt is greater
>than TotalAssets in the Year 2007 or Year 2008.
>         If this is the case, the firm should be dropped out of my data
>completely such that the firm observation is not only dropped for 2007
>and 2008, but also for all the other years.
>         
>        I tried: drop if Year == 2007 | Year == 2008 & LongTermDebt >
>TotalAssets. This made Stata to delete about 5000 observation altough I
>just had 150 observations with LongTermDebt > Total Assets.
>        There is something I do not get as a Stata beginner. Is the drop
>function the wrong function when trying to delete observations under
>several conditions such as comparing Debt to Assets ?
>
>I would be really really happy to receive an answer.
>
>Julian
>*
>*   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/