Bookmark and Share

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: Overwritten?


From   Steve Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Overwritten?
Date   Tue, 31 Dec 2013 12:00:08 -0500

Your problem is easy to diagnose, I think: -sroot.ado contains a
-predict residuals, res- command. This is equivalent to generating a new
variable. At the second iteration, -predict- should fail because the
data set already contains a variable named "res". If I am correct, the
problem has nothing to do with the -if- error that Sergiy's -srootfix-
is designed to work around.

In such situations, I use a -keep if-, -append-, and -save, replace-
approach. Below is the outline of a do file for you to fill in. You
should be able to call -sroot- directly, as -srootfix- is no longer
needed.


**********CODE BEGINS****************
/* Create a data set to hold results */
clear
save dhold, emptyok replace

forval i = 1/`limit' {
use yourdata, clear
keep if id ==`i'
/* Do something that generates variables  */
append using dhold
save dhold, replace
}
**************CODE ENDS**************


Steve

Steven J Samuels
sjsamuels@gmail.com
18 Cantine's Island
Saugerties NY 12477 USA
tel: 845-246-0774


On Dec 30, 2013, at 9:28 PM, Francis, Richard N wrote:

Nick,

Thank you for your advice.

In your experience, is there a reasonable expectation that the original author will alter the program?

Rick 

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
Sent: Monday, December 30, 2013 7:06 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Overwritten?

<>
Side details first:

A convention suggested for Statalist is to use notation such as -foo- to flag names of commands, variables, etc. as used in Stata. That way, for example, the word

replace

is to be understood as just ordinary English but the word

-replace-

is easy to spot as something that could be typed in Stata.

A convention is just that, but I like the Statalist convention, and have often commended it. The different convention you're introducing here of using -FOO instead does not strike me as a good one, if only for the reason that upper case is usually not idiomatic Stata, but that's a personal view.

For the record, -sroot- here is from Stata Journal 9(3) 2009.

The nub of the matter I think is this. The program -srootfix-  by Sergiy Radyakin  -preserve-s your original data, does some work and when it's done -restore-s the original data. Thus it's futile adding new variables to the dataset after -preserve-, as they will just disappear at the -restore-. The -restore- is not explicit in the code, but it's automatic given the -preserve-. The fact that -preserve- requires a -restore- for the fix you originally asked for is explicit in http://www.stata.com/statalist/archive/2013-11/msg00103.html

Short of fixing a clone of -sroot- properly so that you don't need a work-around, or asking the original author to do that for you, it seems that you would need to -save- the dataset you create each time and then -append- them all.

Fixing this properly is the better solution, as then your overall problem should yield to -statsby-.

Nick
njcoxstata@gmail.com


On 30 December 2013 21:59, Francis, Richard N <rnfrancis@utep.edu> wrote:

> Have an option within a called program -SROOTFIX which generates residuals for a regression command.
> 
> However, the variable containing the residuals -RES is absent.
> 
> The called program -SROOTFIX actually calls another program -SROOT.
> 
> Separate execution of -SROOT retains the residuals variable -RES, which suggests that something within the -SROOTFIX pgm overwrites the residuals variable -RES.
> 
> The master program is as follows:
> 
> forval i = 1/`limit' {
>         display "Iteration of `i' of `limit"
>         replace which = "`: label (id) `i''" in `i'
>         srootfix fcf if id == `i' , residuals(res)
>                 qui foreach v in Z14 Z24 Z34 Z4t Z3t Z2t Z1t {
>                  replace `v' = r(`v') in `i'
>         }
> }
> 
> I have executed the program with the -replace command.
> 
> However, the residuals variable -RES continues to be absent.
> 
> The contents of -SROOTFIX are as follows:
> 
> *! By Sergiy Radyakin, 2013
> *! Fixes problem of -if- modifier [suspected] not working in public (SJ) version of -sroot-.
> ** See http://www.stata.com/statalist/archive/2013-11/msg00366.html
> 
> program define srootfix
>   version 9.2
>   syntax varname(ts) [if] [in] ,[*]
>   preserve
>   marksample touse
>   keep if `touse'
>   sroot `varlist' , `options'
> end
> 
> ** end of file **
> 
> Any ideas for the absence of the residuals variable -RES are greatly 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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index