Documentation issue: `est.att.avg` does not exist, is it rather `est.avg`

[MatthieuStigler]

The documentation of gsynth mentions est.att.avg=inference for att.avg. but the slot does not exist!?

It seems it is instead called est.avg it is called?

gsynth/R/core.R

Line 2868 in 114fdce

est.avg <- t(as.matrix(c(att.avg, att.avg.j$se, att.avg.j$CI.l, att.avg.j$CI.u, att.avg.j$P)))

library(gsynth)
#> ## Syntax has been updated since v.1.2.0.
#> ## Comments and suggestions -> yiqingxu@stanford.edu.
data(gsynth)
out <- gsynth(Y ~ D + X1 + X2, data = simdata, parallel = FALSE, 
              index = c("id","time"), force = "two-way",
              CV = TRUE, r = c(0, 2), se = TRUE, inference ="parametric") 
#> Cross-validating ... 
#>  r = 0; sigma2 = 1.84865; IC = 1.02023; PC = 1.74458; MSPE = 2.37280
#>  r = 1; sigma2 = 1.51541; IC = 1.20588; PC = 1.99818; MSPE = 1.71743
#>  r = 2; sigma2 = 0.99737; IC = 1.16130; PC = 1.69046; MSPE = 1.14540
#> 
#>  r* = 2
#> 
#> Simulating errors .....Bootstrapping ...
#> ..
out$est.att.avg
#> NULL
out$est.avg
#>         Estimate     S.E. CI.lower CI.upper p.value
#> ATT.avg 5.543534 0.260916 5.032148  6.05492       0

^{Created on 2024-09-30 with reprex v2.1.1}

[MatthieuStigler]

it seems also Y.tr.cnt and Y.ct.cnt are mentioned in the doc, but not present in the data?

library(gsynth)
#> ## Syntax has been updated since v.1.2.0.
#> ## Comments and suggestions -> yiqingxu@stanford.edu.
data(gsynth)
out <- gsynth(Y ~ D + X1 + X2, data = simdata, parallel = FALSE, 
              index = c("id","time"), force = "two-way",
              CV = TRUE, r = c(0, 5), se = FALSE) 
#> Cross-validating ... 
#>  r = 0; sigma2 = 1.84865; IC = 1.02023; PC = 1.74458; MSPE = 2.37280
#>  r = 1; sigma2 = 1.51541; IC = 1.20588; PC = 1.99818; MSPE = 1.71743
#>  r = 2; sigma2 = 0.99737; IC = 1.16130; PC = 1.69046; MSPE = 1.14540*
#>  r = 3; sigma2 = 0.94664; IC = 1.47216; PC = 1.96215; MSPE = 1.15032
#>  r = 4; sigma2 = 0.89411; IC = 1.76745; PC = 2.19241; MSPE = 1.21397
#>  r = 5; sigma2 = 0.85060; IC = 2.05928; PC = 2.40964; MSPE = 1.23876
#> 
#>  r* = 2
"Y.tr.cnt" %in% names(out)
#> [1] FALSE
"Y.ct.cnt" %in% names(out)
#> [1] FALSE

^{Created on 2024-10-07 with reprex v2.1.1}

[xuyiqing]

Dear Matthieu, Does your data require recentering? Could you attached a panelView plot so that it is easier to diagnose? We're also migrating the package to "fect". We will make sure this get addressed in the new version.

…

---------- Forwarded message --------- From: Matthieu ***@***.***> Date: Mon, Oct 7, 2024 at 4:14 AM Subject: Re: [xuyiqing/gsynth] Documentation issue: `est.att.avg` does not exist, is it rather `est.avg` (Issue #92) To: xuyiqing/gsynth ***@***.***> Cc: Subscribed ***@***.***> it seems also Y.tr.cnt and Y.ct.cnt are mentioned in the doc, but not present in the data? library(gsynth)#> ## Syntax has been updated since v.1.2.0.#> ## Comments and suggestions -> ***@***.*** data(gsynth)out <- gsynth(Y ~ D + X1 + X2, data = simdata, parallel = FALSE, index = c("id","time"), force = "two-way", CV = TRUE, r = c(0, 5), se = FALSE) #> Cross-validating ... #> r = 0; sigma2 = 1.84865; IC = 1.02023; PC = 1.74458; MSPE = 2.37280#> r = 1; sigma2 = 1.51541; IC = 1.20588; PC = 1.99818; MSPE = 1.71743#> r = 2; sigma2 = 0.99737; IC = 1.16130; PC = 1.69046; MSPE = 1.14540*#> r = 3; sigma2 = 0.94664; IC = 1.47216; PC = 1.96215; MSPE = 1.15032#> r = 4; sigma2 = 0.89411; IC = 1.76745; PC = 2.19241; MSPE = 1.21397#> r = 5; sigma2 = 0.85060; IC = 2.05928; PC = 2.40964; MSPE = 1.23876#> #> r* = 2"Y.tr.cnt" %in% names(out)#> [1] FALSE"Y.ct.cnt" %in% names(out)#> [1] FALSE Created on 2024-10-07 with reprex v2.1.1 <https://reprex.tidyverse.org> — Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB2PKGBDCCXQ3AAEFMYK7N3Z2JULDAVCNFSM6AAAAABPCZSK6GVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJWGYZTCNJSGU> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Yiqing Xu Assistant Professor Department of Political Science Stanford University https://yiqingxu.org/

[MatthieuStigler]

Hi

This is just a reprex using the default simdata, so should be easy to diagnose?

Ok, are you saying Y.tr.cnt is only present when the data is centered/has X? If so, might be worth mentioning it in the doc?

Thank a lot!

[MatthieuStigler]

also, interesting to hear about fect. Do you have a page discussing the migration (i.e. advantages/differences of fect vs gsynth, timeline of migration, etc)? If fect is mature, it might be good to ope an issue on gsynth redicreting users to fect? Thanks!

[harrisonhxy]

Hi

This is just a reprex using the default simdata, so should be easy to diagnose?

Ok, are you saying Y.tr.cnt is only present when the data is centered/has X? If so, might be worth mentioning it in the doc?

Thank a lot!

Hello, I also found that gsynth does not have Y.tr.cnt and Y.ct.cnt, and perhaps eff.cnt is also missing. This issue seems to still exist as of v1.3.1 (2026)!
This makes it difficult for me to effectively obtain all the information returned by gsynth. Because as far as I know, Y.tr.cnt and Y.ct.cnt are crucial when drawing Treated and Counterfactual Outcomes (staggered) plots.
So, I would like to ask you, for the situation where Y.tr.cnt and Y.ct.cnt are missing, do you have any good solutions?

[xuyiqing]

Thanks! Could you try "fect" <https://yiqingxu.org/packages/fect/index.html> instead? I might be a change in "fect" that causes this.

…

On Wed, Jan 14, 2026 at 2:30 AM harrisonhxy ***@***.***> wrote: *harrisonhxy* left a comment (xuyiqing/gsynth#92) <#92 (comment)> Hi This is just a reprex using the default simdata, so should be easy to diagnose? Ok, are you saying Y.tr.cnt is only present when the data is centered/has X? If so, might be worth mentioning it in the doc? Thank a lot! Hello, I also found that gsynth does not have Y.tr.cnt and Y.ct.cnt, and perhaps eff.cnt is also missing. This issue seems to still exist as of v1.3.1 (2026)! This makes it difficult for me to effectively obtain all the information returned by gsynth. Because as far as I know, Y.tr.cnt and Y.ct.cnt are crucial when drawing Treated and Counterfactual Outcomes (staggered) plots. So, I would like to ask you, for the situation where Y.tr.cnt and Y.ct.cnt are missing, do you have any good solutions? — Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB2PKGHF4VEIQ5UB4J4U2QL4GYLGHAVCNFSM6AAAAACRU34RGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTONBYHA3DAOJUG4> . You are receiving this because you commented.Message ID: ***@***.***>

-- Yiqing Xu Assistant Professor Department of Political Science Stanford University https://yiqingxu.org/

[TianzhuQin]

Hi
This is just a reprex using the default simdata, so should be easy to diagnose?
Ok, are you saying Y.tr.cnt is only present when the data is centered/has X? If so, might be worth mentioning it in the doc?
Thank a lot!

Hello, I also found that gsynth does not have Y.tr.cnt and Y.ct.cnt, and perhaps eff.cnt is also missing. This issue seems to still exist as of v1.3.1 (2026)! This makes it difficult for me to effectively obtain all the information returned by gsynth. Because as far as I know, Y.tr.cnt and Y.ct.cnt are crucial when drawing Treated and Counterfactual Outcomes (staggered) plots. So, I would like to ask you, for the situation where Y.tr.cnt and Y.ct.cnt are missing, do you have any good solutions?

hi, would you mind send the code and the data to tianzhu@stanford.edu please? I can try to help and look into this!

[harrisonhxy]

Thank you for your reply! The data I use is the data of trunout and the sample code. It is a staggered data structure. I checked the return values of gsynth. They include the following, and there is no Y.tr.cnt; Y.ct.cnt.

>>> The return value of the R package gsynth includes the following variables:
['Y.dat', 'D.dat', 'I.dat', 'Y', 'D', 'X', 'W', 'T.on', 'G', 'balance.period', 'hasRevs', 'T.off', 'T.on.balance', 'index', 'id', 'rawtime', 'binary', 'loo', 'proportion', 'pre.periods', 'tost.threshold', 'placeboTest', 'placebo.period', 'carryoverTest', 'carryover.period', 'unit.type', 'obs.missing', 'obs.missing.balance', 'method', 'Y.ct', 'Y.ct.full', 'Y', 'D', 'tr', 'co', 'eff', 'eff.tr', 'I', 'II', 'att.avg', 'att.avg.unit', 'force', 'T', 'N', 'Ntr', 'Nco', 'p', 'r.cv', 'IC', 'beta', 'est', 'mu', 'niter', 'validX', 'validF', 'time', 'att', 'count', 'eff.calendar', 'N.calendar', 'eff.calendar.fit', 'eff.pre', 'eff.pre.equiv', 'pre.sd', 'PC', 'sigma2', 'sigma2.fect', 'res', 'res.full', 'rmse', 'factor', 'lambda.co', 'lambda.tr', 'wgt.implied', 'alpha.tr', 'alpha.co', 'xi', 'est.avg', 'att.bound', 'att.avg.boot', 'est.avg.unit', 'att.avg.unit.boot', 'est.eff.calendar', 'est.eff.calendar.fit', 'est.att', 'est.att90', 'att.boot', 'att.boot.original', 'att.vcov', 'att.count.boot', 'vartype', 'eff.boot', 'D.boot', 'I.boot', 'colnames.boot', 'beta.boot', 'est.beta', 'test.out', 'call', 'formula', 'data']

[xuyiqing]

We will definitely address this. This is a reverse compatibility issue -- we updated fect but forgot to check gsynth. Thanks for letting us know!

…

On Tue, Jan 20, 2026 at 9:38 PM harrisonhxy ***@***.***> wrote: *harrisonhxy* left a comment (xuyiqing/gsynth#92) <#92 (comment)> Thank you for your reply! The data I use is the data of trunout and the sample code. It is a staggered data structure. I checked the return values of gsynth. They include the following, and there is no Y.tr.cnt; Y.ct.cnt. >>> The return value of the R package gsynth includes the following variables: ['Y.dat', 'D.dat', 'I.dat', 'Y', 'D', 'X', 'W', 'T.on', 'G', 'balance.period', 'hasRevs', 'T.off', 'T.on.balance', 'index', 'id', 'rawtime', 'binary', 'loo', 'proportion', 'pre.periods', 'tost.threshold', 'placeboTest', 'placebo.period', 'carryoverTest', 'carryover.period', 'unit.type', 'obs.missing', 'obs.missing.balance', 'method', 'Y.ct', 'Y.ct.full', 'Y', 'D', 'tr', 'co', 'eff', 'eff.tr', 'I', 'II', 'att.avg', 'att.avg.unit', 'force', 'T', 'N', 'Ntr', 'Nco', 'p', 'r.cv', 'IC', 'beta', 'est', 'mu', 'niter', 'validX', 'validF', 'time', 'att', 'count', 'eff.calendar', 'N.calendar', 'eff.calendar.fit', 'eff.pre', 'eff.pre.equiv', 'pre.sd', 'PC', 'sigma2', 'sigma2.fect', 'res', 'res.full', 'rmse', 'factor', 'lambda.co', 'lambda.tr', 'wgt.implied', 'alpha.tr', 'alpha.co', 'xi', 'est.avg', 'att.bound', 'att.avg.boot', 'est.avg.unit', 'att.avg.unit.boot', 'est.eff.calendar', 'est.eff.calendar.fit', 'est.att', 'est.att90', 'att.boot', 'att.boot.original', 'att.vcov', 'att.count.boot', 'vartype', 'eff.boot', 'D.boot', 'I.boot', 'colnames.boot', 'beta.boot', 'est.beta', 'test.out', 'call', 'formula', 'data'] — Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB2PKGEMBFKZRRQHJVOKOF34H4GGJAVCNFSM6AAAAACRU34RGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTONZWGI4DMNJSGQ> . You are receiving this because you commented.Message ID: ***@***.***>

-- Yiqing Xu Assistant Professor Department of Political Science Stanford University https://yiqingxu.org/