<>
Seems like the combination btw " id==1 " and " estimation_window==1 " simply
does not exist. As I said, put -capture- in front of the -regress- command
(and any other command that might choke on missing combinations). The fact
mentioned in your initial post that you replaced all missings with a dot
does not change this story, BTW.
HTH
Martin
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Dias Marios
Sent: Samstag, 4. Juli 2009 10:26
To: [email protected]
Subject: Re: st: AW: no observations R 2000 error
Hi,
seems that id is running ok from 1 to 85. I have put -set trace on- but i am
afraid i don't understand much. Seems like some versions are missing but i
have no idea what these are. What does version stand for?
Thank you
THIS IS THE OUTPUT I GET
. set trace on
. set more off
.
. gen predicted_return=.
(13686 missing values generated)
.
. egen id=group(company_id)
----------------------------------------------------------------------------
----------- begin egen ---
- version 6, missing
- gettoken type 0 : 0, parse(" =(")
- gettoken name 0 : 0, parse(" =(")
- if `"`name'"'=="=" {
= if `"="'=="=" {
- local name `"`type'"'
= local name `"id"'
- local type : set type
- }
- else {
gettoken eqsign 0 : 0, parse(" =(")
if `"`eqsign'"' != "=" {
error 198
}
}
- confirm new variable `name'
= confirm new variable id
- gettoken fcn 0 : 0, parse(" =(")
- gettoken args 0 : 0, parse(" ,") match(par)
- if `"`par'"' != "(" {
= if `"("' != "(" {
exit 198
}
- if `"`args'"' == "_all" {
= if `"company_id"' == "_all" {
version 7.0, missing
unab args : _all
local args : subinstr local args "`_sortindex'" "", all word
version 6.0, missing
}
- syntax [if] [in] [, *]
- if _by() {
local byopt "by(`_byvars')"
local cma ","
}
- else if `"`options'"' != "" {
= else if `""' != "" {
local cma ","
}
- tempvar dummy
- global EGEN_Varname `name'
= global EGEN_Varname id
- version 7.0, missing
- global EGEN_SVarname `_sortindex'
= global EGEN_SVarname __000000
- version 6.0, missing
- capture noisily _g`fcn' `type' `dummy' = (`args') `if' `in' `cma' `byopt'
`options'
= capture noisily _ggroup float __000001 = (company_id)
----------------------------------------------------------------------------
---- begin _ggroup ---
- version 7, missing
- gettoken type 0 : 0
- gettoken g 0 : 0
- gettoken eqs 0 : 0
- syntax varlist [if] [in] [, Missing BY(string) Label LName(name)
Truncate(numlist max=1 int >= 1
> )]
- if `"`by'"' != "" {
= if `""' != "" {
_egennoby group() `"`by'"'
}
- if "`truncate'" != "" & "`label'" == "" {
= if "" != "" & "" == "" {
di as err "truncate() option requires the label option"
exit 198
}
- if "`lname'" != "" {
= if "" != "" {
local label "label"
}
- tempvar touse
- quietly {
- mark `touse' `if' `in'
= mark __000002
- if "`missing'"=="" {
= if ""=="" {
- markout `touse' `varlist', strok
= markout __000002 company_id, strok
- }
- sort `touse' `varlist'
= sort __000002 company_id
- quietly by `touse' `varlist': gen `type' `g'=1 if _n==1 & `touse'
= quietly by __000002 company_id: gen float __000001=1 if _n==1 & __000002
- replace `g'=sum(`g')
= replace __000001=sum(__000001)
- replace `g'=. if `touse'!=1
= replace __000001=. if __000002!=1
- if "`label'"!="" {
= if ""!="" {
if "`lname'" == "" {
local lname $EGEN_Varname
unab lab_name_list : _all
foreach var of local lab_name_list {
local lab : value label `var'
local lab_list `lab_list' `lab'
}
local dup_check : list lname in lab_list
if `dup_check' {
di as err "you must specify the lname() option"
exit 198
}
}
local dfltfmt : set dp
local dfltfmt = cond("`dfltfmt'"=="period","%9.0g","%9,0g")
count if !`touse'
local j = 1 + r(N)
sum `g', meanonly
local max `r(max)'
local i 1
while `i' <= 0`max' {
tokenize `varlist'
local vtmp " "
local x 1
while "`1'"!="" {
local vallab : value label `1'
local val = `1'[`j']
if "`vallab'" != "" {
local vtmp2 : label `vallab' `val' `truncate'
}
else {
cap confirm numeric var `1'
if _rc==0 {
local vtmp2 = string(`1'[`j'],"`dfltfmt'")
}
else {
if "`truncate'"=="" {
local vtmp2 = trim(`1'[`j'])
}
else {
local vtmp2 = trim(substr(trim(`1'[`j']),1,`truncate'))
}
}
}
local x = `x' + length("`vtmp2'") + 1
local vtmp "`vtmp' `vtmp2'"
mac shift
}
local val `vtmp'
label def `lname' `i' "`val'", modify
count if `g' == `i'
local j = `j' + r(N)
local i = `i' + 1
}
label val `g' `lname'
}
- }
- if length("group(`varlist')") > 80 {
= if length("group(company_id)") > 80 {
note `g' : group(`varlist')
label var `g' "see notes"
}
- else label var `g' "group(`varlist')"
= else label var __000001 "group(company_id)"
----------------------------------------------------------------------------
------ end _ggroup ---
- global EGEN_SVarname
- global EGEN_Varname
- if _rc { exit _rc }
- quietly count if missing(`dummy')
= quietly count if missing(__000001)
- if r(N) {
local s = cond(r(N)>1,"s","")
di in bl "(" r(N) " missing value`s' generated)"
}
- rename `dummy' `name'
= rename __000001 id
----------------------------------------------------------------------------
----------- end egen ---
.
.
.
. forvalues i=1(1)85 {
2.
. l id company_id if id==`i' & dif==0
3.
. reg ret market_return if id==`i' & estimation_window==1
4.
. predict p if id==`i'
5.
. replace predicted_return = p if id==`i' & event_window==1
6.
. drop p
7.
. }
- forvalues i=1(1)85 {
- l id company_id if id==`i' & dif==0
= l id company_id if id==1 & dif==0
+---------------+
| id compan~d |
|---------------|
121. | 1 1 |
+---------------+
- reg ret market_return if id==`i' & estimation_window==1
= reg ret market_return if id==1 & estimation_window==1
----------------------------------------------------------------------------
---------- begin reg ---
- if _by() {
local by "by `_byvars'`_byrc0':"
}
- `by' regress `0'
= regress ret market_return if id==1 & estimation_window==1
----------------------------------------------------------------------------
---- begin regress ---
- version 9, missing
- local version : di "version " string(_caller()) ", missing :"
- if _by() {
local BY `"by `_byvars'`_byrc0':"'
}
- syntax [anything] [if] [in] [aw fw iw pw] [, VCE(passthru) Robust
CLuster(passthru) * ]
- if `:length local vce' {
= if 0 {
`version' `BY' _vce_parserun regress, mark(CLuster) : `0'
if "`s(exit)'" != "" {
ereturn local cmdline `"regress `0'"'
exit
}
_vce_parse, argopt(CLuster) opt(OLS Robust HC2 HC3) old :
[`weight'`exp'], `vce' `robust' `clust
> er'
if "`r(cluster)'" != "" {
local cluster cluster(`r(cluster)')
}
else if "`r(vce)'" != "ols" {
local robust = "`r(vce)'"
}
local vce = cond("`r(vce)'" != "", "`r(vce)'", "ols")
}
- else if "`cluster'" != "" {
= else if "" != "" {
local vce cluster
}
- else if "`weight'" == "pweight" | "`robust'" != "" {
= else if "" == "pweight" | "" != "" {
local vce robust
}
- else local vce ols
- if replay() {
if _by() {
error 190
}
`version' Display `0'
exit
}
- `version' `BY' _regress `anything' `if' `in' [`weight'`exp'], `options'
`robust' `cluster'
= version 10, missing : _regress ret market_return if id==1 &
estimation_window==1 [],
no observations
----------------------------------------------------------------------------
------ end regress ---
----------------------------------------------------------------------------
------------ end reg ---
predict p if id==`i'
replace predicted_return = p if id==`i' & event_window==1
drop p
}
r(2000);
. edit
- preserve
--- On Fri, 7/3/09, Martin Weiss <[email protected]> wrote:
> From: Martin Weiss <[email protected]>
> Subject: st: AW: no observations R 2000 error
> To: [email protected]
> Date: Friday, July 3, 2009, 8:14 AM
>
> <>
>
> Could be several things, but it is critically important for
> the list to know
> after which line Stata complains. Also, you could -set
> trace on- and report
> the area around the error.
>
> Are you sure you have id running from 1 to 85? Type
> -levelsof id- to check!
> If one of them is missing, the -reg- command throws up an
> error and stops
> the whole thing... You can prevent this by adding -capture-
> in front of the
> command in question...
>
>
>
> HTH
> Martin
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]]
> Im Auftrag von Dias Marios
> Gesendet: Freitag, 3. Juli 2009 08:02
> An: [email protected]
> Betreff: st: no observations R 2000 error
>
> Hi
>
> I know this type of error has been posted before but i have
> followed some
> suggested solutions and i still cannot find a way to solve
> it.
>
> THIS IS MY LIST OF VARIABLES
>
> target_name
> str34 %34s
> company_id
> byte %8.0g
> market_return
> float %9.0g
> ret
> float %9.0g
> date
> float %td
> event_date
> float %td
> datenum
> float %9.0g
> td
> float %9.0g
> dif
> float %9.0g
> event_window float
> %9.0g
> count_event_obs float
> %9.0g
> estimation_window float
> %9.0g
>
> AND HERE IS THE CODE I AM TRYING TO RUN AND GET THE ERROR
>
> set more off
>
> gen predicted_return=.
> egen id=group(company_id)
> forvalues i=1(1)85 {
> l id company_id if id==`i' & dif==0
> reg ret market_return if id==`i' &
> estimation_window==1
> predict p if id==`i'
> replace predicted_return = p if id==`i'
> & event_window==1
> drop p
> }
>
>
> I have checked that in my variables all missing values are
> replaced with a "
> . " so i would guess that is not the problem.
>
> Thanks
>
> Marios
>
>
>
>
> *
> * 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/
>
>
> *
> * 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/
>
*
* 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/
*
* 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/
