Re: st: Dealing with panel data

[Nick Cox <njcoxstata@gmail.com>]

Without rancour, and a definite grin on my face, I don't feel
responsible for not suggesting solutions to questions that weren't
asked, which is what your post ends by asking about.

This is an example where you understand what you want but it's hard
for anyone else to read all your words and absorb them all and then
emit Stata code. I think you would be better off posting sample data
and sample results for what you want and then ask how to do it. Make
it, naturally, as simple as possible.

Meanwhile the paper that I cited is full of tricks in this neighbourhood.

Nick
njcoxstata@gmail.com


On 7 August 2013 16:31, John Kenny <john.kenny27@mail.dcu.ie> wrote:
> Thanks Nick,
>
> Having looked over your the solution you recommended it does not quite
> achieve what I am looking for. The previous solution got the mean of
> the overround for every race held at a particular meeting. The first
> command found the number of times the favorite won the first race at
> the each meeting. The second command I as I understand it gets the
> mean of the overround when 'fav_won_1 & race_no > 1'  for each
> meeting.
>
> 1) egen fav_won_1 = total(favourite == 1 & race_no == 1), by(meeting)
> 2) egen mean_overround = mean(overround/(fav_won_1 & race_no > 1)), by(meeting)
>
> This however is not what I am trying to analyse and I apologise for
> the poor explanation of my problem. The overall solution I want to get
> is the mean of the overround for each race depending on the outcome of
> the previous races at that meeting on a particular day .  Examples of
> this are get the mean of the overround for the 4TH race if the
> favorite won the first race, another example is getting the mean of
> the overround for the second race if the favorite won the first race,
> along with other variations of races and whether the favorite won
> either the 1st, 2nd, or 3rd race.
>
> Any Further help on this would be greatly appreciated.
>
> Thanks
> John
>
> On 6 August 2013 16:54, Nick Cox <njcoxstata@gmail.com> wrote:
>> You need to let the other observations in each group know what
>> happened in race 1. That sounds like
>>
>> egen fav_won_1 = total(favourite == 1 & race_no == 1), by(meeting)
>>
>> which will give you a meeting-wide indicator variable.
>>
>> Then you go
>>
>> egen mean_overround = mean(overround/(fav_won_1 & race_no > 1)), by(meeting)
>>
>> The division sign / is not a typo. Fortuitously, but fortunately, it
>> works here as if it were a condition sign |.
>>
>> For a more systematic review of such tricks, see
>>
>> http://www.stata-journal.com/article.html?article=dm0055
>>
>> P.S. -summarize- is a command, not a function.
>> Nick
>> njcoxstata@gmail.com
>>
>>
>> On 6 August 2013 16:44, John Kenny <john.kenny27@mail.dcu.ie> wrote:
>>> Dear Statalist,
>>>
>>> I'm relatively new to stata and I cannot find a standard way too solve
>>> my problem and I may need to write a .*do file.
>>>
>>> I'm dealing with a very large data set that has about 20 variable that
>>> outlines horse racing results with 711,000 observations. The problem
>>> that I am having is that I cannot get the mean of one variable
>>> 'overround' if two other variables are a certain value.
>>>
>>>  To be more specific within the data set I am using the 'overround'
>>> determines the bookmakers profit margin. What I want to do is get the
>>> mean of the 'overround'  for each race from 2 to the last race 7 if
>>> the 'favorite' (which is a dummy variable) won the first race. I have
>>> a number of variables that outline the time of each race and when it
>>> occurs at a certain meeting. These are some of the variables that are
>>> outlined in the data set for each race there is a variable that says
>>> were the race is held [ 'Meeting' ], the date and time, ['date',
>>> 'time'] , the odds given for each horse ['odds'], the race number at
>>> that meeting ['race_no' (lists the races from 1-7)], whether the
>>> favorite won that race ['favorite'] and the overround which is the
>>> bookmakers profit [ 'overround' ].
>>>
>>> What I have tried is using the summarize function and try and get the
>>> mean of the 'overround' if 'race_no'==1 & 'favorite'==1. However every
>>> combination of variables I tried using it always just got the mean of
>>> the 'overround' for race 1 if the favourite won and not the mean of
>>> the second race or third race if the favorite won the first race.
>>>
>>> Any help would be greatly appreciated as I have been stuck on this for a while.
>>>
>>> Thanks in advance.
>>>
>>> John
>>> *
>>> *   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/