Please ignore, I discovered my error
↧
reg with -if var==1- or -inlist- omits more observations than satisfy the criteria
↧
Diff-in-diff: Can a pre-treatment placebo test be used to control for the anticipatory effect?
Hi.
I am examining whether a policy change impacted return on assets. The policy was announced in 2002 but implemented in 2005. Can you examine the anticipatory effect by performing a diff-in-diff using 2002 as the placebo date?
Would including a lead or lag in my baseline regression aid the detection of a potential anticipatory effect?
Many thanks!
I am examining whether a policy change impacted return on assets. The policy was announced in 2002 but implemented in 2005. Can you examine the anticipatory effect by performing a diff-in-diff using 2002 as the placebo date?
Would including a lead or lag in my baseline regression aid the detection of a potential anticipatory effect?
Many thanks!
↧
↧
With -bayesmh-, what is the proper syntax for a custom prior density for a factor variable's parameter that is a function of the parameter?
With bayesmh, you can create a custom prior density with the density() option of prior().
And that probability density function can be a function of the parameter (see first example below).
And a parameter for a factor variable can have a custom density (see second example below).
But when I try to create a custom density that combines both of these, that is, is a function of the parameter for a factor variable, I get an error message whose complaints I've ruled out (third example below).
Here's the code (I have also attached a do-file containing the complete code for two reproducible examples)
and here's the result (I've also attached an SMCL log file that contains the result).
.ÿ/*ÿNotÿaÿproblemÿforÿtheÿcustomÿdensityÿtoÿbeÿaÿfunctionÿofÿaÿparameterÿ(orÿitsÿlinearÿcombi
>ÿnation)ÿ*/
.ÿbayesmhÿforeign,ÿlikelihood(logistic)ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:_cons},ÿdensity(exp({foreign:_cons})ÿ/ÿ(1ÿ+ÿexp({foreign:_cons}))^2))
>ÿÿ///
>ÿÿÿÿÿÿÿÿÿnomodelsummary
ÿÿ
Burn-inÿ...
Simulationÿ...
BayesianÿlogisticÿregressionÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿiterationsÿÿ=ÿÿÿÿÿ12,500
Random-walkÿMetropolis-HastingsÿsamplingÿÿÿÿÿÿÿÿÿBurn-inÿÿÿÿÿÿÿÿÿÿ=ÿÿÿÿÿÿ2,500
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿsampleÿsizeÿ=ÿÿÿÿÿ10,000
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿNumberÿofÿobsÿÿÿÿ=ÿÿÿÿÿÿÿÿÿ74
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿAcceptanceÿrateÿÿ=ÿÿÿÿÿÿ.4486
Logÿmarginalÿlikelihoodÿ=ÿ-47.010554ÿÿÿÿÿÿÿÿÿÿÿÿÿEfficiencyÿÿÿÿÿÿÿ=ÿÿÿÿÿÿ.2163
ÿ
------------------------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿÿÿÿÿ|ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿEqual-tailed
ÿÿÿÿÿforeignÿ|ÿÿÿÿÿÿMeanÿÿÿStd.ÿDev.ÿÿÿÿÿMCSEÿÿÿÿÿMedianÿÿ[95%ÿCred.ÿInterval]
-------------+----------------------------------------------------------------
ÿÿÿÿÿÿÿ_consÿ|ÿ-.8417671ÿÿÿ.2630011ÿÿÿ.005654ÿÿ-.8390754ÿÿ-1.384746ÿÿ-.3504125
------------------------------------------------------------------------------
.ÿ
.ÿ/*ÿNotÿaÿproblemÿforÿaÿfactorÿvariableÿtoÿhaveÿcustomÿdensityÿpriorÿ*/
.ÿbayesmhÿforeignÿi.k,ÿlikelihood(logistic)ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:_cons},ÿnormal(0,ÿ5))ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:i1.k},ÿdensity(1))ÿ///
>ÿÿÿÿÿÿÿÿÿnomodelsummary
ÿÿ
Burn-inÿ...
Simulationÿ...
BayesianÿlogisticÿregressionÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿiterationsÿÿ=ÿÿÿÿÿ12,500
Random-walkÿMetropolis-HastingsÿsamplingÿÿÿÿÿÿÿÿÿBurn-inÿÿÿÿÿÿÿÿÿÿ=ÿÿÿÿÿÿ2,500
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿsampleÿsizeÿ=ÿÿÿÿÿ10,000
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿNumberÿofÿobsÿÿÿÿ=ÿÿÿÿÿÿÿÿÿ74
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿAcceptanceÿrateÿÿ=ÿÿÿÿÿÿ.2666
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿEfficiency:ÿÿminÿ=ÿÿÿÿÿÿÿ.136
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿavgÿ=ÿÿÿÿÿÿ.1436
Logÿmarginalÿlikelihoodÿ=ÿ-47.085067ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿmaxÿ=ÿÿÿÿÿÿ.1513
ÿ
------------------------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿÿÿÿÿ|ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿEqual-tailed
ÿÿÿÿÿforeignÿ|ÿÿÿÿÿÿMeanÿÿÿStd.ÿDev.ÿÿÿÿÿMCSEÿÿÿÿÿMedianÿÿ[95%ÿCred.ÿInterval]
-------------+----------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿ1.kÿ|ÿ-.0355851ÿÿÿ.4936124ÿÿÿÿ.01269ÿÿ-.0221173ÿÿ-.9684661ÿÿÿ.9078394
ÿÿÿÿÿÿÿ_consÿ|ÿ-.8504277ÿÿÿ.3452926ÿÿÿ.009364ÿÿ-.8448572ÿÿÿ-1.54843ÿÿÿ-.212137
------------------------------------------------------------------------------
.ÿ
.ÿ/*ÿProblemÿwhenÿcustomÿdensityÿpriorÿisÿaÿfunctionÿofÿtheÿfactorÿvariable'sÿ
>ÿÿÿÿparameterÿ(orÿitsÿlinearÿcombination)ÿ*/
.ÿbayesmhÿforeignÿi.k,ÿlikelihood(logistic)ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:_cons},ÿnormal(0,ÿ5))ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:i1.k},ÿdensity(exp({foreign:i1.k})ÿ/ÿ(1ÿ+ÿexp({foreign:i1.k}))^2))
noÿpriorÿspecifiedÿforÿforeign:i1.k
ÿÿÿÿPriorÿdistributionsÿmustÿbeÿspecifiedÿforÿallÿmodelÿparameters.ÿYouÿmayÿhaveÿomitted
ÿÿÿÿoptionÿprior()ÿforÿsomeÿofÿtheÿparametersÿorÿmistypedÿtheÿnamesÿofÿtheÿparametersÿinÿthe
ÿÿÿÿspecifiedÿprior()ÿoptions.
r(198);
And that probability density function can be a function of the parameter (see first example below).
And a parameter for a factor variable can have a custom density (see second example below).
But when I try to create a custom density that combines both of these, that is, is a function of the parameter for a factor variable, I get an error message whose complaints I've ruled out (third example below).
Here's the code (I have also attached a do-file containing the complete code for two reproducible examples)
Code:
quietly sysuse auto
generate byte k = mod(_n, 2)
/* Not a problem for the custom density to be a function of a parameter (or its linear combination) */
bayesmh foreign, likelihood(logistic) ///
prior({foreign:_cons}, density(exp({foreign:_cons}) / (1 + exp({foreign:_cons}))^2)) ///
nomodelsummary
/* Not a problem for a factor variable to have custom density prior */
bayesmh foreign i.k, likelihood(logistic) ///
prior({foreign:_cons}, normal(0, 5)) ///
prior({foreign:i1.k}, density(1)) ///
nomodelsummary
/* Problem when custom density prior is a function of the factor variable's
parameter (or its linear combination) */
bayesmh foreign i.k, likelihood(logistic) ///
prior({foreign:_cons}, normal(0, 5)) ///
prior({foreign:i1.k}, density(exp({foreign:i1.k}) / (1 + exp({foreign:i1.k}))^2))
exit
.ÿ/*ÿNotÿaÿproblemÿforÿtheÿcustomÿdensityÿtoÿbeÿaÿfunctionÿofÿaÿparameterÿ(orÿitsÿlinearÿcombi
>ÿnation)ÿ*/
.ÿbayesmhÿforeign,ÿlikelihood(logistic)ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:_cons},ÿdensity(exp({foreign:_cons})ÿ/ÿ(1ÿ+ÿexp({foreign:_cons}))^2))
>ÿÿ///
>ÿÿÿÿÿÿÿÿÿnomodelsummary
ÿÿ
Burn-inÿ...
Simulationÿ...
BayesianÿlogisticÿregressionÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿiterationsÿÿ=ÿÿÿÿÿ12,500
Random-walkÿMetropolis-HastingsÿsamplingÿÿÿÿÿÿÿÿÿBurn-inÿÿÿÿÿÿÿÿÿÿ=ÿÿÿÿÿÿ2,500
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿsampleÿsizeÿ=ÿÿÿÿÿ10,000
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿNumberÿofÿobsÿÿÿÿ=ÿÿÿÿÿÿÿÿÿ74
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿAcceptanceÿrateÿÿ=ÿÿÿÿÿÿ.4486
Logÿmarginalÿlikelihoodÿ=ÿ-47.010554ÿÿÿÿÿÿÿÿÿÿÿÿÿEfficiencyÿÿÿÿÿÿÿ=ÿÿÿÿÿÿ.2163
ÿ
------------------------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿÿÿÿÿ|ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿEqual-tailed
ÿÿÿÿÿforeignÿ|ÿÿÿÿÿÿMeanÿÿÿStd.ÿDev.ÿÿÿÿÿMCSEÿÿÿÿÿMedianÿÿ[95%ÿCred.ÿInterval]
-------------+----------------------------------------------------------------
ÿÿÿÿÿÿÿ_consÿ|ÿ-.8417671ÿÿÿ.2630011ÿÿÿ.005654ÿÿ-.8390754ÿÿ-1.384746ÿÿ-.3504125
------------------------------------------------------------------------------
.ÿ
.ÿ/*ÿNotÿaÿproblemÿforÿaÿfactorÿvariableÿtoÿhaveÿcustomÿdensityÿpriorÿ*/
.ÿbayesmhÿforeignÿi.k,ÿlikelihood(logistic)ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:_cons},ÿnormal(0,ÿ5))ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:i1.k},ÿdensity(1))ÿ///
>ÿÿÿÿÿÿÿÿÿnomodelsummary
ÿÿ
Burn-inÿ...
Simulationÿ...
BayesianÿlogisticÿregressionÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿiterationsÿÿ=ÿÿÿÿÿ12,500
Random-walkÿMetropolis-HastingsÿsamplingÿÿÿÿÿÿÿÿÿBurn-inÿÿÿÿÿÿÿÿÿÿ=ÿÿÿÿÿÿ2,500
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿMCMCÿsampleÿsizeÿ=ÿÿÿÿÿ10,000
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿNumberÿofÿobsÿÿÿÿ=ÿÿÿÿÿÿÿÿÿ74
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿAcceptanceÿrateÿÿ=ÿÿÿÿÿÿ.2666
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿEfficiency:ÿÿminÿ=ÿÿÿÿÿÿÿ.136
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿavgÿ=ÿÿÿÿÿÿ.1436
Logÿmarginalÿlikelihoodÿ=ÿ-47.085067ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿmaxÿ=ÿÿÿÿÿÿ.1513
ÿ
------------------------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿÿÿÿÿ|ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿEqual-tailed
ÿÿÿÿÿforeignÿ|ÿÿÿÿÿÿMeanÿÿÿStd.ÿDev.ÿÿÿÿÿMCSEÿÿÿÿÿMedianÿÿ[95%ÿCred.ÿInterval]
-------------+----------------------------------------------------------------
ÿÿÿÿÿÿÿÿÿ1.kÿ|ÿ-.0355851ÿÿÿ.4936124ÿÿÿÿ.01269ÿÿ-.0221173ÿÿ-.9684661ÿÿÿ.9078394
ÿÿÿÿÿÿÿ_consÿ|ÿ-.8504277ÿÿÿ.3452926ÿÿÿ.009364ÿÿ-.8448572ÿÿÿ-1.54843ÿÿÿ-.212137
------------------------------------------------------------------------------
.ÿ
.ÿ/*ÿProblemÿwhenÿcustomÿdensityÿpriorÿisÿaÿfunctionÿofÿtheÿfactorÿvariable'sÿ
>ÿÿÿÿparameterÿ(orÿitsÿlinearÿcombination)ÿ*/
.ÿbayesmhÿforeignÿi.k,ÿlikelihood(logistic)ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:_cons},ÿnormal(0,ÿ5))ÿ///
>ÿÿÿÿÿÿÿÿÿprior({foreign:i1.k},ÿdensity(exp({foreign:i1.k})ÿ/ÿ(1ÿ+ÿexp({foreign:i1.k}))^2))
noÿpriorÿspecifiedÿforÿforeign:i1.k
ÿÿÿÿPriorÿdistributionsÿmustÿbeÿspecifiedÿforÿallÿmodelÿparameters.ÿYouÿmayÿhaveÿomitted
ÿÿÿÿoptionÿprior()ÿforÿsomeÿofÿtheÿparametersÿorÿmistypedÿtheÿnamesÿofÿtheÿparametersÿinÿthe
ÿÿÿÿspecifiedÿprior()ÿoptions.
r(198);
↧
virtual joins
I noticed JMP 13 and Spotfire 7.5 have virtual joins or similar for multiple data tables. So you don't have to physically join or merge data tables to pull data together with a unique identifier. Does Stata have this capability? Should it? It is supposed to save memory and make data changes faster in large data sets in some situations.
↧
Stata 15 release
Good morning,
There were a lot of reports (and excitement!) that Stata 15 was going to be released this week, e.g. www.econjobrumors.com/topic/stata-15-will-be-released-on-tuesday, and many, many, more. Obviously this did not happen, so a lot of users are understandably very anxious. Does StataCorp have an update on the release date?
Thanks, E.
There were a lot of reports (and excitement!) that Stata 15 was going to be released this week, e.g. www.econjobrumors.com/topic/stata-15-will-be-released-on-tuesday, and many, many, more. Obviously this did not happen, so a lot of users are understandably very anxious. Does StataCorp have an update on the release date?
Thanks, E.
↧
↧
Clustering Diff-in-Diff
Hi.
I am working on a diff-in-diff analysis on a policy intervention on wage gaps in two different sectors. I know that the sector and time-specific random effects could generate a possible clustering problem, which may ruin the inference of my analysis. I only have two clusters in the two sectors. However, what would be the consequence of cluster the standard errors on an individual level?
Thank you.
Fredrik
I am working on a diff-in-diff analysis on a policy intervention on wage gaps in two different sectors. I know that the sector and time-specific random effects could generate a possible clustering problem, which may ruin the inference of my analysis. I only have two clusters in the two sectors. However, what would be the consequence of cluster the standard errors on an individual level?
Thank you.
Fredrik
↧
Saving Tempfiles
Hi there,
Using the code below, I've appended all Excel files in my "ImportantFiles" folder to a tempfile called datasets.
How do I save the content in the tempfile as a .dta file in C:/Documents?
Thanks!
Erika
Using the code below, I've appended all Excel files in my "ImportantFiles" folder to a tempfile called datasets.
Code:
cd C:/Documents/ImportantFiles
local myfiles : dir . files "*.xlsx"
display `"`myfiles'"'
tempfile datasets
foreach file of local myfiles {
import excel `"`file'"', firstrow clear
capture append using "`datasets'"
save "`datasets'", replace
}
Thanks!
Erika
↧
Differential effects of a continuous variable
I am looking at the effect of private schooling on partisanship and have found an interesting result: Some private schooling makes individuals appear to be decidedly more likely than those with no private school, but the liberalizing effect slows dramatically after 4 years of private school and actually reverses after 6 years of private school. To illustrate:
Array
My guess is that the "liberalizing" effect is being driven by students who attend private elementary schools. I do have variables telling me whether the individual was in a public or private school at each grade level, so this should be testable. Is there a more practical way other than creating dummies to test my hypothesis?
Array
My guess is that the "liberalizing" effect is being driven by students who attend private elementary schools. I do have variables telling me whether the individual was in a public or private school at each grade level, so this should be testable. Is there a more practical way other than creating dummies to test my hypothesis?
↧
generate quantile regression graphs for dummy variable
Hi all
I am trying to generate quantile regression coefficients graph using grqreg command. I run the following commands: (Note: y is my dependent variable, gender, x1, x2, are independent variables. gender is dummy variable which assumes value 1 for female and 0 otherwise.)
qui sqreg y i.gender x1 x2 x3, quantile(.05 .10 .25 .50 .75 .90.95)
grqreg, ols olsci
After I run these commands, I get the following message
1b.gender invalid name
r(198);
Is it that this command does not accommodate the graphs for dummy variables, or I am doing any mistake?
Kindly help
Regards
Karim
I am trying to generate quantile regression coefficients graph using grqreg command. I run the following commands: (Note: y is my dependent variable, gender, x1, x2, are independent variables. gender is dummy variable which assumes value 1 for female and 0 otherwise.)
qui sqreg y i.gender x1 x2 x3, quantile(.05 .10 .25 .50 .75 .90.95)
grqreg, ols olsci
After I run these commands, I get the following message
1b.gender invalid name
r(198);
Is it that this command does not accommodate the graphs for dummy variables, or I am doing any mistake?
Kindly help
Regards
Karim
↧
↧
spmap error
I am looking at how a specific policy diffuses through states. I put together a very basic data set with 50 observations, one for each state, and no missing data. It's in wide for with a binary variable for whether the policy exists for each year (e.g., p2009, p2010... p2014). I am trying to create an individual map for each of these to show which states had already adopted the policy in each year. I am running the following code:
spmap p2009 using uscoord if statename!="Alaska" & statename!="Hawaii", id(id) fcolor(Blues)
for years 2009 through 2014. Each one brings up the map with the correct data, except for 2011 and 2014. On these two years I receive the following error:
When no attribute variable is specified, option fcolor() does not accept palette names
r(198);
Even when I drop the fcolor specification (e.g., spmap p2011 using uscoord if statename!="Alaska" & statename!="Hawaii", id(id)), I get the exact same error. There doesn't appear to be anything different about the data in these two years, so I cannot figure out why only 2 of the 6 maps won't show up. I'm new to spmap, so any advice would help. I didn't see any similar issues in other spmap posts, so I hope this is not a duplicate question.
Thanks,
James
spmap p2009 using uscoord if statename!="Alaska" & statename!="Hawaii", id(id) fcolor(Blues)
for years 2009 through 2014. Each one brings up the map with the correct data, except for 2011 and 2014. On these two years I receive the following error:
When no attribute variable is specified, option fcolor() does not accept palette names
r(198);
Even when I drop the fcolor specification (e.g., spmap p2011 using uscoord if statename!="Alaska" & statename!="Hawaii", id(id)), I get the exact same error. There doesn't appear to be anything different about the data in these two years, so I cannot figure out why only 2 of the 6 maps won't show up. I'm new to spmap, so any advice would help. I didn't see any similar issues in other spmap posts, so I hope this is not a duplicate question.
Thanks,
James
↧
Multiplying or dividing coefficients to create risk score
Greetings statalist,
I'm planning to do a similar analyses found in: Balkus, Jennifer E., et al. "An empiric HIV risk scoring tool to predict HIV-1 acquisition in African women." JAIDS Journal of Acquired Immune Deficiency Syndromes 72.3 (2016): 333-343. (https://pdfs.semanticscholar.org/54e...30ffe9e5d5.pdf). Taken from the paper:
"To identify the combination of factors that best predicted HIV risk, we used forward and backward stepwise Cox proportional hazards model that evaluated the inclusion or exclusion of potential predictors at each step. All models were stratified by study site. The model with the lowest Akaike information criterion was chosen as the final model for the risk score. Individual predictors included in the final model were assigned a score by dividing the coefficient for the predictor in the final model by the lowest coefficient among all predictors in the model and rounding to the nearest integer. The sum of the values for each predictor represented the total score for each participant, and the HIV incidence for each total score category was calculated. The predictive ability of the total score and each predictor was assessed by calculating area under the receiver operating characteristic curve. The score was internally validated using 10-fold cross- validation, and the area under the curve (AUC) for the final model was compared with the mean AUC of the 10 different models. Additional performance characteristics (sensitivity, specificity, positive predictive value, and negative predictive value) were calculated using risk score cut-points that corresponded to an HIV incidence in the risk score category of approximately >3% and >5%. Incidence curves were generated to assess cumulative HIV incidence by risk score cut-point."
Conceptually, I understand what was done. Unfortunately, I do not know how to execute the analyses after the regression models. I've read similar papers that multiplied the regression coefficient by 10 and then round to the nearest integer. I would greatly appreciate assistance in identifying the STATA commands associated with executing this: "Individual predictors included in the final model were assigned a score by dividing the coefficient for the predictor in the final model by the lowest coefficient among all predictors in the model and rounding to the nearest integer." as well as "The score was internally validated using 10-fold cross- validation, and the area under the curve (AUC) for the final model was compared with the mean AUC of the 10 different models."
Thanks so much for your assistance.
I'm planning to do a similar analyses found in: Balkus, Jennifer E., et al. "An empiric HIV risk scoring tool to predict HIV-1 acquisition in African women." JAIDS Journal of Acquired Immune Deficiency Syndromes 72.3 (2016): 333-343. (https://pdfs.semanticscholar.org/54e...30ffe9e5d5.pdf). Taken from the paper:
"To identify the combination of factors that best predicted HIV risk, we used forward and backward stepwise Cox proportional hazards model that evaluated the inclusion or exclusion of potential predictors at each step. All models were stratified by study site. The model with the lowest Akaike information criterion was chosen as the final model for the risk score. Individual predictors included in the final model were assigned a score by dividing the coefficient for the predictor in the final model by the lowest coefficient among all predictors in the model and rounding to the nearest integer. The sum of the values for each predictor represented the total score for each participant, and the HIV incidence for each total score category was calculated. The predictive ability of the total score and each predictor was assessed by calculating area under the receiver operating characteristic curve. The score was internally validated using 10-fold cross- validation, and the area under the curve (AUC) for the final model was compared with the mean AUC of the 10 different models. Additional performance characteristics (sensitivity, specificity, positive predictive value, and negative predictive value) were calculated using risk score cut-points that corresponded to an HIV incidence in the risk score category of approximately >3% and >5%. Incidence curves were generated to assess cumulative HIV incidence by risk score cut-point."
Conceptually, I understand what was done. Unfortunately, I do not know how to execute the analyses after the regression models. I've read similar papers that multiplied the regression coefficient by 10 and then round to the nearest integer. I would greatly appreciate assistance in identifying the STATA commands associated with executing this: "Individual predictors included in the final model were assigned a score by dividing the coefficient for the predictor in the final model by the lowest coefficient among all predictors in the model and rounding to the nearest integer." as well as "The score was internally validated using 10-fold cross- validation, and the area under the curve (AUC) for the final model was compared with the mean AUC of the 10 different models."
Thanks so much for your assistance.
↧
Testing Fama-Macbeth; Shanken-Correction
Hi all,
I made some asset pricing models with Fama-Macbeth method and I would like to apply the Shanken-correction but how could I possibly add this to my code?
Thanks,
Thomas Jansen
I made some asset pricing models with Fama-Macbeth method and I would like to apply the Shanken-correction but how could I possibly add this to my code?
Code:
quietly {
foreach var of varlist BH1 BL1 SH1 SL1 BH2 BL2 SH2 SL2 BH3 BL3 SH3 SL3 BH4 BL4 SH4 SL4 BH5 BL5 SH5 SL5 BH6 BL6 SH6 SL6 BH7 BL7 SH7 SL7 {
replace `var'=`var'-Rf
}
}
** Number of observations defining **
local obs=14802
tempname betas
postfile `betas' beta_c beta_HML_Factor using "D:\betas.dta", replace
quietly {
foreach y of varlist BH1 BL1 SH1 SL1 BH2 BL2 SH2 SL2 BH3 BL3 SH3 SL3 BH4 BL4 SH4 SL4 BH5 BL5 SH5 SL5 BH6 BL6 SH6 SL6 BH7 BL7 SH7 SL7 {
regress `y' HML_Factor if fyear>=1962 & fyear<=2017
scalar beta_c=_b[_cons]
scalar beta_HML_Factor=_b[HML_Factor]
post `betas' (beta_c) (beta_HML_Factor)
}
}
postclose `betas'
drop HML_Factor
xpose, clear varname
drop in 1
merge 1:1 _n using "D:\betas.dta"
tempname lambdas
local obs=14802
postfile `lambdas' lambda_c lambda_HML_Factor lambda_r2 using "D:\lambdas.dta", replace
quietly {
forvalues i=1(1)`obs' {
regress v`i' beta_HML_Factor
scalar lambda_c=_b[_cons]
scalar lambda_HML_Factor=_b[beta_HML_Factor]
scalar lambda_r2=e(r2)
post `lambdas' (lambda_c) (lambda_HML_Factor)(lambda_r2)
}
}
postclose `lambdas'
snapshot restore 1
use "D:\lambdas.dta", clear
local obs=14802
quietly {
foreach var of varlist (lambda_c) (lambda_HML_Factor) {
summarize `var'
scalar `var'_mean=r(mean)
scalar `var'_tratio=sqrt(`obs')*r(mean)/r(sd)
}
summarize lambda_r2
scalar lambda_r2_mean=r(mean)
}
display lambda_c_mean
display lambda_c_tratio
display lambda_HML_Factor_mean
display lambda_HML_Factor_tratio
display lambda lambda_r2_mean
Thanks,
Thomas Jansen
↧
Goodness-of-fit tests and variable selection for a zero-inflated negative binomial model
Hello - I am working on an analysis of over-dispersed count data using zero-inflated binomial regression and am having difficulty figuring out the appropriate goodness-of-fit test(s) to use and selecting the parameters for the model. Here is an overview of my data:
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------------------
phys_hlth | 1,727 4.10249 8.11829 0 30
Because of the size of the variance relative to the mean, I moved from a ZIP model to ZINB. Putting some of the predictors mentioned above into the ZINB model returns good numbers for overall chi2, alpha, and the Vuong test, as well as for my main predictor (DOV_LGBT):
. zinb phys_hlth DOV_LGBT discrim PPAGE, inflate (PPAGE DOV_LGBT) vuong zip
Zero-inflated negative binomial regression Number of obs = 1,554
Nonzero obs = 634
Zero obs = 920
Inflation model = logit LR chi2(3) = 12.45
Log likelihood = -3113.399 Prob > chi2 = 0.0060
------------------------------------------------------------------------------
phys_hlth | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
phys_hlth |
DOV_LGBT | .2159001 .096089 2.25 0.025 .0275691 .404231
discrim | .2479829 .1694438 1.46 0.143 -.0841209 .5800867
PPAGE | .0087431 .0030843 2.83 0.005 .0026981 .0147882
_cons | 1.592725 .1875341 8.49 0.000 1.225165 1.960286
-------------+----------------------------------------------------------------
inflate |
PPAGE | .0126378 .0039695 3.18 0.001 .0048578 .0204179
DOV_LGBT | -.4053969 .1249087 -3.25 0.001 -.6502134 -.1605805
_cons | -.4047859 .2509455 -1.61 0.107 -.8966301 .0870582
-------------+----------------------------------------------------------------
/lnalpha | .2658189 .104051 2.55 0.011 .0618826 .4697553
-------------+----------------------------------------------------------------
alpha | 1.304499 .1357345 1.063837 1.599603
------------------------------------------------------------------------------
Likelihood-ratio test of alpha=0: chibar2(01) = 3673.90 Pr>=chibar2 = 0.0000
Vuong test of zinb vs. standard negative binomial: z = 6.47 Pr>z = 0.0000
What I am struggling with is the following:
1) Are there other goodness-of-fit tests that I should be running to ensure that the ZINB model is a good fit? I have used the margins command to estimate the predicted means after doing a robust ZINB regression, and these estimates are close to the actual means (even without the analytical weight), but I want to make sure I'm not stumbling blindly into the ZINB model because I can't think of any other approaches:
. margins DOV_LGBT
Predictive margins Number of obs = 1,554
Model VCE : Robust
Expression : Predicted number of events, predict()
------------------------------------------------------------------------------
| Delta-method
| Margin Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
DOV_LGBT |
0 | 3.432679 .2549347 13.46 0.000 2.933016 3.932342
LGBT | 5.237526 .35971 14.56 0.000 4.532507 5.942545
------------------------------------------------------------------------------
Actual means:
. mean phys_hlth [aw=weight_1], over(DOV_LGBT)
Mean estimation Number of obs = 1,727
_subpop_1: DOV_LGBT = 0
LGBT: DOV_LGBT = LGBT
--------------------------------------------------------------
Over | Mean Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
phys_hlth |
_subpop_1 | 3.555177 .2547231 3.055579 4.054776
LGBT | 4.984303 .3125431 4.3713 5.597306
--------------------------------------------------------------
2) Is there a method to selecting which variables to inflate in the ZINB model?
3) Relatedly, what is the best method to use with ZINB regression to select which variables to include in the model at all? E.g., I have taken gender out because it was consistently showing up as nonsignificant no matter whether I inflated it or not, but is testing each of my ~15 possible independent variable one-by-one like that my only option? Can I use something like <gvselect> or forwards/backwards selection with a ZINB model, and if so, how?
Thank you!
- Dependent variable: "phys_hlth," which is the number of days in the last month when respondent's self-reported physical health was not good (0-30)
- Main predictor: "DOV_LGBT," which is LGBT identity (0=not LGBT-identified, 1=LGBT-identified)
- Other possible predictors/controls: age (continuous; "PPAGE" in the commands below), and recent experience of discrimination (binary; "discrim" in the commands below), as well as others such as race (5 categories), gender (binary), insurance status (binary), etc.
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------------------
phys_hlth | 1,727 4.10249 8.11829 0 30
Because of the size of the variance relative to the mean, I moved from a ZIP model to ZINB. Putting some of the predictors mentioned above into the ZINB model returns good numbers for overall chi2, alpha, and the Vuong test, as well as for my main predictor (DOV_LGBT):
. zinb phys_hlth DOV_LGBT discrim PPAGE, inflate (PPAGE DOV_LGBT) vuong zip
Zero-inflated negative binomial regression Number of obs = 1,554
Nonzero obs = 634
Zero obs = 920
Inflation model = logit LR chi2(3) = 12.45
Log likelihood = -3113.399 Prob > chi2 = 0.0060
------------------------------------------------------------------------------
phys_hlth | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
phys_hlth |
DOV_LGBT | .2159001 .096089 2.25 0.025 .0275691 .404231
discrim | .2479829 .1694438 1.46 0.143 -.0841209 .5800867
PPAGE | .0087431 .0030843 2.83 0.005 .0026981 .0147882
_cons | 1.592725 .1875341 8.49 0.000 1.225165 1.960286
-------------+----------------------------------------------------------------
inflate |
PPAGE | .0126378 .0039695 3.18 0.001 .0048578 .0204179
DOV_LGBT | -.4053969 .1249087 -3.25 0.001 -.6502134 -.1605805
_cons | -.4047859 .2509455 -1.61 0.107 -.8966301 .0870582
-------------+----------------------------------------------------------------
/lnalpha | .2658189 .104051 2.55 0.011 .0618826 .4697553
-------------+----------------------------------------------------------------
alpha | 1.304499 .1357345 1.063837 1.599603
------------------------------------------------------------------------------
Likelihood-ratio test of alpha=0: chibar2(01) = 3673.90 Pr>=chibar2 = 0.0000
Vuong test of zinb vs. standard negative binomial: z = 6.47 Pr>z = 0.0000
What I am struggling with is the following:
1) Are there other goodness-of-fit tests that I should be running to ensure that the ZINB model is a good fit? I have used the margins command to estimate the predicted means after doing a robust ZINB regression, and these estimates are close to the actual means (even without the analytical weight), but I want to make sure I'm not stumbling blindly into the ZINB model because I can't think of any other approaches:
. margins DOV_LGBT
Predictive margins Number of obs = 1,554
Model VCE : Robust
Expression : Predicted number of events, predict()
------------------------------------------------------------------------------
| Delta-method
| Margin Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
DOV_LGBT |
0 | 3.432679 .2549347 13.46 0.000 2.933016 3.932342
LGBT | 5.237526 .35971 14.56 0.000 4.532507 5.942545
------------------------------------------------------------------------------
Actual means:
. mean phys_hlth [aw=weight_1], over(DOV_LGBT)
Mean estimation Number of obs = 1,727
_subpop_1: DOV_LGBT = 0
LGBT: DOV_LGBT = LGBT
--------------------------------------------------------------
Over | Mean Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
phys_hlth |
_subpop_1 | 3.555177 .2547231 3.055579 4.054776
LGBT | 4.984303 .3125431 4.3713 5.597306
--------------------------------------------------------------
2) Is there a method to selecting which variables to inflate in the ZINB model?
3) Relatedly, what is the best method to use with ZINB regression to select which variables to include in the model at all? E.g., I have taken gender out because it was consistently showing up as nonsignificant no matter whether I inflated it or not, but is testing each of my ~15 possible independent variable one-by-one like that my only option? Can I use something like <gvselect> or forwards/backwards selection with a ZINB model, and if so, how?
Thank you!
↧
↧
Padding names - string match
Hey all,
I had an earlier post http://www.statalist.org/forums/foru...-large-dataset
in which one of the lister shared a portion of his code (#3). He suggested padding the names before doing name cleaning ,
suppose the content of name (shortname)
would this only insert a blank before First, and one blank after Corp?
or would it insert the blanks before and after for each word ?
A more general question, if our dataset is commercial database which include firm names, is it always necessary to pad the names before we do a task such as
thanks,
Rochelle
I had an earlier post http://www.statalist.org/forums/foru...-large-dataset
in which one of the lister shared a portion of his code (#3). He suggested padding the names before doing name cleaning ,
Code:
* Make sure that we have a space at the beginning and end so that we can
* search words delimited by spaces.
replace shortname = " " + shortname + " "
listdiff shortname, reset
replace shortname = subinstr(shortname," INCORPORATED ", " ",.)
Code:
First Eagle Corp
or would it insert the blanks before and after for each word ?
A more general question, if our dataset is commercial database which include firm names, is it always necessary to pad the names before we do a task such as
Code:
replace shortname = subinstr(shortname," INCORPORATED ", " ",.)
Rochelle
↧
Stuck on effect sizes
Hi all!
I have two independent groups who went through the same treatment. I want to see the effect size. I know the following code is often used:
esize twosample math, by(grouptype) cohensd hedgesg glassdelta
However, there is a problem. The study has a pre-test/post-test set up. For the pre-test scores, one group was significantly higher than the other. Therefore, it seems like looking at the post-test effect size isn't helpful. It might be more helpful to find an effect size for the differences in scores. However, for the life of me I can't figure out how to do this on Stata. I'd love any advice. Thanks in advance for the help!
I have two independent groups who went through the same treatment. I want to see the effect size. I know the following code is often used:
esize twosample math, by(grouptype) cohensd hedgesg glassdelta
However, there is a problem. The study has a pre-test/post-test set up. For the pre-test scores, one group was significantly higher than the other. Therefore, it seems like looking at the post-test effect size isn't helpful. It might be more helpful to find an effect size for the differences in scores. However, for the life of me I can't figure out how to do this on Stata. I'd love any advice. Thanks in advance for the help!
↧
Looking for advice on computing the difference between two variables with different scales to compute a third variable
Hello
I was hoping someone could help me with the below query, if so it would be greatly appreciated:
All the best
Conal
I was hoping someone could help me with the below query, if so it would be greatly appreciated:
- I have two variables with separate scales/metrics: (i) self-report extroversion (possible score of 1-5); (ii) friends report of extroversion (possible score of 1-8)
- I would like to calculate the meaningful difference between "self-report extroversion" and "friends report of extroversion" for each participant in my sample but I need the two variables to be on the same scale to do this. (I would like to create a new variable based on this difference called "extroversion congruence").
- Would anybody be able to advise me on the best way to handle the above in stata?
- I hope this makes sense but if not I am happy to clarify
All the best
Conal
↧
Help: Graph Supply and Demand from Regression
Reg. With IV.do
Hey!
I'm pretty new with stata and have been trying lately to obtain the demand and supply of natural gas for the US (short and long term)
I have been successful with my regressions so far (log(Ct)=log (Ct-1) + log (Pt)+Weather+date+ui, with absorbed fixed effects (by month and region), robust error and using an instrumental variable for the price). Also, my regressions are separated between winter and summer elasticities (with dummies) for consumption to find the difference in winter and summer demand.
areg log_con_total hat_log_spot_price winter_hat lag_log_con_total w_lag_log_con_total hdd date, absorb(region_x_mes) robust
areg log_prod_total hat_log_spot_price winter_hat w_lag_log_prod_total lag_log_prod_total date hdd, absorb (region_x_mes) robust
(hat_log_spot_price is the IV, w_ implies the winter dummy, hdd is for weather)
Now I want to graph the supply and demand for those regressions, but do not know how (how do I get the supply and demand equations from results?); I especially want to work with short term supply and demand (so I don't care about long term elasticities) and separate winter and summer demand.
Any comments might help,
Thanks for your time
![]( "Smile")
Hey!
I'm pretty new with stata and have been trying lately to obtain the demand and supply of natural gas for the US (short and long term)
I have been successful with my regressions so far (log(Ct)=log (Ct-1) + log (Pt)+Weather+date+ui, with absorbed fixed effects (by month and region), robust error and using an instrumental variable for the price). Also, my regressions are separated between winter and summer elasticities (with dummies) for consumption to find the difference in winter and summer demand.
areg log_con_total hat_log_spot_price winter_hat lag_log_con_total w_lag_log_con_total hdd date, absorb(region_x_mes) robust
areg log_prod_total hat_log_spot_price winter_hat w_lag_log_prod_total lag_log_prod_total date hdd, absorb (region_x_mes) robust
(hat_log_spot_price is the IV, w_ implies the winter dummy, hdd is for weather)
Now I want to graph the supply and demand for those regressions, but do not know how (how do I get the supply and demand equations from results?); I especially want to work with short term supply and demand (so I don't care about long term elasticities) and separate winter and summer demand.
Any comments might help,
Thanks for your time
↧
↧
Order in Bar Graphs
I am having difficulty putting my bars in a specific order. It automatically orders the bars from left to right alphabetically but I want each president in chronological order. How do I do this?My current code is this:
graph bar laborforce, over(president) ytitle("Unemployment Rate") title("Average Monthly Unemployment Rate") subtitle("for each U.S. President")
graph bar laborforce, over(president) ytitle("Unemployment Rate") title("Average Monthly Unemployment Rate") subtitle("for each U.S. President")
↧
Data Cleaning
I have a variable service_number. I know it HAS TO have 10 characters. How can I generate an indicator for all the observations for which service_number != 10. Also, how can I delete these observations in one command?
↧
Recursive regression with fixed training window
Dear All,
Once again I am encountering some specific regression which cannot correctly execute. The idea is, lets say we have time series sample 1930-2010, we split this sample at 1980 and use prior observations as a fixed window, t=49, {1,...,t} and run regression on some variable lets say. At year 1980, we start running recursive regressions and using this window as base, thus now we run regression with {1,..,t+1} and so on until we exhaust all the remaining time series, until 2010, i.e {1,..,t+1},..,{1,..,t+30}. Maybe someone has some ideas or had similar issue?
Thank you in advance!
Marijus
Once again I am encountering some specific regression which cannot correctly execute. The idea is, lets say we have time series sample 1930-2010, we split this sample at 1980 and use prior observations as a fixed window, t=49, {1,...,t} and run regression on some variable lets say. At year 1980, we start running recursive regressions and using this window as base, thus now we run regression with {1,..,t+1} and so on until we exhaust all the remaining time series, until 2010, i.e {1,..,t+1},..,{1,..,t+30}. Maybe someone has some ideas or had similar issue?
Thank you in advance!
Marijus
↧