[Bug]: rdrobust v11.1.0

[gianmarcovignes]

Implementation

Stata

Summary

I have Stata 19.5 on Windows, using rdrobust v11.1.0 with clustered standard errors it always triggers me this error, which I can't solve: : 3499 rdrobust_vce_qq_cluster() not found
Mata run-time error

Minimal Example

sysuse auto, clear
rdrobust price mpg, vce(cluster rep78)

Expected Behavior

No response

Environment

No response

[mdcattaneo]

Thanks for reporting this. I could not reproduce the Mata error with the v11.1.0 files currently in the repository in a fresh Stata 16.1 session.

There is one correction needed in the minimal example: mpg does not straddle the default cutoff c(0), so the example should be

sysuse auto, clear
rdrobust price mpg, c(20) vce(cluster rep78)

With that correction, the command completes successfully and reports CR1 standard errors.

The function named in the error, rdrobust_vce_qq_cluster(), is included in the distributed lrdrobust.mlib. This error therefore suggests that Stata is finding the v11.1.0 rdrobust.ado file but not the matching/current Mata library, which can occur after a partial installation or an upgrade in an already-running Stata session.

Could you please close and reopen Stata, then run the following and post the output?

discard
mata: mata mlib index
mata: mata which rdrobust_vce_qq_cluster()
which rdrobust
findfile lrdrobust.mlib
about
adopath

Then please retry the corrected example above. If mata which does not report lrdrobust, reinstalling the complete package (rather than replacing only rdrobust.ado) should resolve the mismatch. The requested output will also let us determine whether there is a Stata 19.5-specific issue.

[gianmarcovignes]

Hi, thank you very much for your answer. I'm sorry for my late response but I was busy working on ohter things. Anyway, this morning I needed again to use rdrobust command, with the function rdrobust_vce_qq_cluster(), and I retried to use it after what you suggested me, but it still doesn't work. Reinstalling rdrobust didn't fix anything, the output you requested me to produce was before reinstalling, and is also after, the following:
discard

.
. mata: mata mlib index
.mlib libraries to be searched are now
lmatabase;lmataado;lmatabma;lmatacate;lmatacollect;lmatadata;lmataerm;lmatafc;lmatagsem;

lmatah2o;lmatahetdid;lmataivqreg;lmatajm;lmatalasso;lmatamcmc;lmatameta;lmatami;lmatamixlo
g;lmatanumlib;lmataopt;lmatapath;lmatapostest;lmatapss;lmatarw;lmatasem;lmatasp;lmatasubex
pr;lmatasvy;lmatatab;lmatats

.
. mata: mata which rdrobust_vce_qq_cluster()
function rdrobust_vce_qq_cluster() not found
r(111);

.
. which rdrobust
C:\Users\giamm\ado\plus\r\rdrobust.ado
*! version 11.1.0 22may2026

.
. findfile lrdrobust.mlib
file "lrdrobust.mlib" not found
r(601);

.
. about

StataNow/SE 19.5 for Windows (64-bit x86-64)
Revision 08 Apr 2025
Copyright 1985-2025 StataCorp LLC

Total physical memory: 16.00 GB
Available physical memory: 6.54 GB

Stata license: Unlimited-user network, expiring 30 Aug 2026
Serial number: 401909302651
Licensed to: Universita' Bocconi Milano
Universita' Bocconi Milano

.
. adopath
[1] (BASE) "C:\Program Files\StataNow19\ado\base/"
[2] (SITE) "C:\Program Files\StataNow19\ado\site/"
[3] "."
[4] (PERSONAL) "C:\Users\giamm\ado\personal/"
[5] (PLUS) "C:\Users\giamm\ado\plus/"
[6] (OLDPLACE) "c:\ado/"

[mdcattaneo]

Thanks—your diagnostic output identifies the problem. This is a packaging error in the SSC distribution, not a Stata 19.5-specific issue. SSC is currently installing rdrobust.ado v11.1.0 without the matching Mata library containing rdrobust_vce_qq_cluster(). Reinstalling from SSC therefore does not resolve it.

As an immediate workaround, please uninstall the SSC copy and install the complete package directly from GitHub:

ado uninstall rdrobust
net install rdrobust, from(https://raw.githubusercontent.com/rdpackages/rdrobust/main/stata) replace

Then restart Stata and verify:

mata: mata mlib index
findfile lrdrobust.mlib
mata: mata which rdrobust_vce_qq_cluster()

sysuse auto, clear
rdrobust price mpg, c(20) vce(cluster rep78)

We will also correct the SSC distribution. Thank you for providing the diagnostic output.