LR Confidence interval for unobs_tr despite vcov matrix is PD

[yushuf]

data.zip

library(transtat)
pdata <- read.csv("pdata.csv")
xdata <- read.csv("xdata.csv")
coef <- read.csv("coef.csv", header = TRUE)

pairwise data to be combined with external data

pdat <- pdata
pdat$interval <- pdat$exit - pdat$entry
pdat$external <- 0
pdat$trace <- 1
pdat$start <- 0
pdat$stop <- pdat$interval

external data

xdat <- xdata
names(xdat)[names(xdat) == "x"] <- "susx"
xdat$external <- 1
xdat$infectious <- NA
xdat$infx <- 0
xdat$interval <- xdat$exit # entry is at time 0
xdat$start <- xdat$entry
xdat$stop <- xdat$exit
dat <- rbind(pdat, xdat)

#Transreg works
unobs_tr <- transreg(
Surv(start, stop, infset) ~ infx + susx + ext(external),
sus = "susceptible", data = dat, subset = (trace == 1),
dist = "weibull", xdist = "exponential"
)
#Variance-covariance matrix is PD
eigen(vcov(unobs_tr))$values >.00

#confint keep working for long time, I have waited overnight.
ci_unobs_tr_lr <-confint(unobs_tr, type = 'lr')

#This errors are not informative
confint(unobs_tr, "log(shape)", type = 'lr')
Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") :
missing value where TRUE/FALSE needed

confint(unobs_tr, "(Intercept)", type = 'lr')
Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") :
missing value where TRUE/FALSE needed

[ekenah]

Hi Yushuf, Please tell me what you found in the data. If the simulations are working, you should write a new draft of the paper and send it to me in early January. Thanks, Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:51 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) data.zip<https://github.com/ekenah/transtat/files/1579862/data.zip> library(transtat) pdata <- read.csv("pdata.csv") xdata <- read.csv("xdata.csv") coef <- read.csv("coef.csv", header = TRUE) pairwise data to be combined with external data pdat <- pdata pdat$interval <- pdat$exit - pdat$entry pdat$external <- 0 pdat$trace <- 1 pdat$start <- 0 pdat$stop <- pdat$interval external data xdat <- xdata names(xdat)[names(xdat) == "x"] <- "susx" xdat$external <- 1 xdat$infectious <- NA xdat$infx <- 0 xdat$interval <- xdat$exit # entry is at time 0 xdat$start <- xdat$entry xdat$stop <- xdat$exit dat <- rbind(pdat, xdat) #Transreg works unobs_tr <- transreg( Surv(start, stop, infset) ~ infx + susx + ext(external), sus = "susceptible", data = dat, subset = (trace == 1), dist = "weibull", xdist = "exponential" ) #Variance-covariance matrix is PD eigen(vcov(unobs_tr))$values >.00 #confint keep working for long time, I have waited overnight. ci_unobs_tr_lr <-confint(unobs_tr, type = 'lr') #This errors are not informative confint(unobs_tr, "log(shape)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed confint(unobs_tr, "(Intercept)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#8>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mZKgKG1Ys0BB8YU8Xx80joWSeguBks5tCoyqgaJpZM4RKFLz>.

[yushuf]

I can share share all data except the LR CI for unobs_tr. Let me proceed with that, in the mean time if we find any solution, we will incorporate that into the paper. That is what I have decided. Thanks

…

________________________________ From: Eben Kenah <notifications@github.com> Sent: Thursday, December 21, 2017 11:53:29 AM To: ekenah/transtat Cc: Sharker, Yushuf; Author Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hi Yushuf, Please tell me what you found in the data. If the simulations are working, you should write a new draft of the paper and send it to me in early January. Thanks, Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:51 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) data.zip<https://github.com/ekenah/transtat/files/1579862/data.zip> library(transtat) pdata <- read.csv("pdata.csv") xdata <- read.csv("xdata.csv") coef <- read.csv("coef.csv", header = TRUE) pairwise data to be combined with external data pdat <- pdata pdat$interval <- pdat$exit - pdat$entry pdat$external <- 0 pdat$trace <- 1 pdat$start <- 0 pdat$stop <- pdat$interval external data xdat <- xdata names(xdat)[names(xdat) == "x"] <- "susx" xdat$external <- 1 xdat$infectious <- NA xdat$infx <- 0 xdat$interval <- xdat$exit # entry is at time 0 xdat$start <- xdat$entry xdat$stop <- xdat$exit dat <- rbind(pdat, xdat) #Transreg works unobs_tr <- transreg( Surv(start, stop, infset) ~ infx + susx + ext(external), sus = "susceptible", data = dat, subset = (trace == 1), dist = "weibull", xdist = "exponential" ) #Variance-covariance matrix is PD eigen(vcov(unobs_tr))$values >.00 #confint keep working for long time, I have waited overnight. ci_unobs_tr_lr <-confint(unobs_tr, type = 'lr') #This errors are not informative confint(unobs_tr, "log(shape)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed confint(unobs_tr, "(Intercept)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#8>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mZKgKG1Ys0BB8YU8Xx80joWSeguBks5tCoyqgaJpZM4RKFLz>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ekenah_transtat_issues_8-23issuecomment-2D353400924&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=BVF4DRIGRhK7DP1u7K_48nxzPxZPSdRewh2pQX1EvM0&e=>, or mute the thread<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEkA32lmsmeFZslLXYS5IYUeWS0c2Gvuks5tCo0JgaJpZM4RKFLz&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=eCg8am9Touy2XLkS7Hx1E6YXIfEAsWcliCFOgShKWaM&e=>.

[ekenah]

Hi Yushuf, In your code, I don’t see you doing stochastic annealing (optim_method = “SANN”) first and then Nelder-Mead. You should do that for each simulation. Use the model with Nelder-Mead for the LR confidence intervals. Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:59 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Eben Kenah <kenah.1@osu.edu>, Comment <comment@noreply.github.com> Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) I can share share all data except the LR CI for unobs_tr. Let me proceed with that, in the mean time if we find any solution, we will incorporate that into the paper. That is what I have decided. Thanks

…

________________________________ From: Eben Kenah <notifications@github.com> Sent: Thursday, December 21, 2017 11:53:29 AM To: ekenah/transtat Cc: Sharker, Yushuf; Author Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hi Yushuf, Please tell me what you found in the data. If the simulations are working, you should write a new draft of the paper and send it to me in early January. Thanks, Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:51 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) data.zip<https://github.com/ekenah/transtat/files/1579862/data.zip> library(transtat) pdata <- read.csv("pdata.csv") xdata <- read.csv("xdata.csv") coef <- read.csv("coef.csv", header = TRUE) pairwise data to be combined with external data pdat <- pdata pdat$interval <- pdat$exit - pdat$entry pdat$external <- 0 pdat$trace <- 1 pdat$start <- 0 pdat$stop <- pdat$interval external data xdat <- xdata names(xdat)[names(xdat) == "x"] <- "susx" xdat$external <- 1 xdat$infectious <- NA xdat$infx <- 0 xdat$interval <- xdat$exit # entry is at time 0 xdat$start <- xdat$entry xdat$stop <- xdat$exit dat <- rbind(pdat, xdat) #Transreg works unobs_tr <- transreg( Surv(start, stop, infset) ~ infx + susx + ext(external), sus = "susceptible", data = dat, subset = (trace == 1), dist = "weibull", xdist = "exponential" ) #Variance-covariance matrix is PD eigen(vcov(unobs_tr))$values >.00 #confint keep working for long time, I have waited overnight. ci_unobs_tr_lr <-confint(unobs_tr, type = 'lr') #This errors are not informative confint(unobs_tr, "log(shape)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed confint(unobs_tr, "(Intercept)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#8>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mZKgKG1Ys0BB8YU8Xx80joWSeguBks5tCoyqgaJpZM4RKFLz>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ekenah_transtat_issues_8-23issuecomment-2D353400924&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=BVF4DRIGRhK7DP1u7K_48nxzPxZPSdRewh2pQX1EvM0&e=>, or mute the thread<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEkA32lmsmeFZslLXYS5IYUeWS0c2Gvuks5tCo0JgaJpZM4RKFLz&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=eCg8am9Touy2XLkS7Hx1E6YXIfEAsWcliCFOgShKWaM&e=>. — You are receiving this because you commented. Reply to this email directly, view it on GitHub<#8 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mXCA3UCtuhzrRwYixZetg_KtHE2Xks5tCo5ggaJpZM4RKFLz>.

[yushuf]

Hello Eben, I tried those options, but had the same trouble with the LR CI for unobs_tr. During trial, I never get LR CI with those options with those options. So I wrote the code as, if that fits without the option, I will try first. If not converges, then I will try with those options. if still not converges, then I will skip the result. If converges without options, then with or without the options provides the same estimates. I get the estimates sometimes. But the problem is I don't know when the problem is happening so that I can control. Thanks Yushuf Thanks

…

________________________________ From: Eben Kenah <notifications@github.com> Sent: Thursday, December 21, 2017 3:14 PM To: ekenah/transtat Cc: Sharker, Yushuf; Author Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hi Yushuf, In your code, I don’t see you doing stochastic annealing (optim_method = “SANN”) first and then Nelder-Mead. You should do that for each simulation. Use the model with Nelder-Mead for the LR confidence intervals. Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:59 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Eben Kenah <kenah.1@osu.edu>, Comment <comment@noreply.github.com> Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) I can share share all data except the LR CI for unobs_tr. Let me proceed with that, in the mean time if we find any solution, we will incorporate that into the paper. That is what I have decided. Thanks

________________________________ From: Eben Kenah <notifications@github.com> Sent: Thursday, December 21, 2017 11:53:29 AM To: ekenah/transtat Cc: Sharker, Yushuf; Author Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hi Yushuf, Please tell me what you found in the data. If the simulations are working, you should write a new draft of the paper and send it to me in early January. Thanks, Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:51 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) data.zip<https://github.com/ekenah/transtat/files/1579862/data.zip> library(transtat) pdata <- read.csv("pdata.csv") xdata <- read.csv("xdata.csv") coef <- read.csv("coef.csv", header = TRUE) pairwise data to be combined with external data pdat <- pdata pdat$interval <- pdat$exit - pdat$entry pdat$external <- 0 pdat$trace <- 1 pdat$start <- 0 pdat$stop <- pdat$interval external data xdat <- xdata names(xdat)[names(xdat) == "x"] <- "susx" xdat$external <- 1 xdat$infectious <- NA xdat$infx <- 0 xdat$interval <- xdat$exit # entry is at time 0 xdat$start <- xdat$entry xdat$stop <- xdat$exit dat <- rbind(pdat, xdat) #Transreg works unobs_tr <- transreg( Surv(start, stop, infset) ~ infx + susx + ext(external), sus = "susceptible", data = dat, subset = (trace == 1), dist = "weibull", xdist = "exponential" ) #Variance-covariance matrix is PD eigen(vcov(unobs_tr))$values >.00 #confint keep working for long time, I have waited overnight. ci_unobs_tr_lr <-confint(unobs_tr, type = 'lr') #This errors are not informative confint(unobs_tr, "log(shape)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed confint(unobs_tr, "(Intercept)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#8>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mZKgKG1Ys0BB8YU8Xx80joWSeguBks5tCoyqgaJpZM4RKFLz>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ekenah_transtat_issues_8-23issuecomment-2D353400924&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=BVF4DRIGRhK7DP1u7K_48nxzPxZPSdRewh2pQX1EvM0&e=>, or mute the thread<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEkA32lmsmeFZslLXYS5IYUeWS0c2Gvuks5tCo0JgaJpZM4RKFLz&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=eCg8am9Touy2XLkS7Hx1E6YXIfEAsWcliCFOgShKWaM&e=>. — You are receiving this because you commented. Reply to this email directly, view it on GitHub<#8 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mXCA3UCtuhzrRwYixZetg_KtHE2Xks5tCo5ggaJpZM4RKFLz>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ekenah_transtat_issues_8-23issuecomment-2D353447123&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=Cs4-oDYX8ujQoVeGnHBFeGvH5IBcuklbJP0DS3RZbGs&s=alw5NcEMCDRNAkEeMi241R-_AOBNbmXcMPbyewpBbtM&e=>, or mute the thread<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEkA36ounvnkkEd9BvK0tsHqXX24jANUks5tCrwQgaJpZM4RKFLz&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=Cs4-oDYX8ujQoVeGnHBFeGvH5IBcuklbJP0DS3RZbGs&s=VSDADrm0EDRSK63YNyrc3XXLFBbiqDNFzatHIaXen3A&e=>.

[ekenah]

Hi Yushuf, Let’s proceed as you suggested with everything but the LR confidence intervals for unobs_tr. Everything else is working, right. If you can, please try to figure out what is happening with the LR confidence intervals. One way to start would be to send single jobs to the cluster. If the job takes more than a certain amount of time, have it return the data to you. Finally, let’s keep the discussion of the simulations off GitHub because it doesn’t have to do with an error in the code. Thanks, Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Friday, December 22, 2017 at 9:26 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Eben Kenah <kenah.1@osu.edu>, Comment <comment@noreply.github.com> Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hello Eben, I tried those options, but had the same trouble with the LR CI for unobs_tr. During trial, I never get LR CI with those options with those options. So I wrote the code as, if that fits without the option, I will try first. If not converges, then I will try with those options. if still not converges, then I will skip the result. If converges without options, then with or without the options provides the same estimates. I get the estimates sometimes. But the problem is I don't know when the problem is happening so that I can control. Thanks Yushuf Thanks

…

________________________________ From: Eben Kenah <notifications@github.com> Sent: Thursday, December 21, 2017 3:14 PM To: ekenah/transtat Cc: Sharker, Yushuf; Author Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hi Yushuf, In your code, I don’t see you doing stochastic annealing (optim_method = “SANN”) first and then Nelder-Mead. You should do that for each simulation. Use the model with Nelder-Mead for the LR confidence intervals. Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:59 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Eben Kenah <kenah.1@osu.edu>, Comment <comment@noreply.github.com> Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) I can share share all data except the LR CI for unobs_tr. Let me proceed with that, in the mean time if we find any solution, we will incorporate that into the paper. That is what I have decided. Thanks

________________________________ From: Eben Kenah <notifications@github.com> Sent: Thursday, December 21, 2017 11:53:29 AM To: ekenah/transtat Cc: Sharker, Yushuf; Author Subject: Re: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) Hi Yushuf, Please tell me what you found in the data. If the simulations are working, you should write a new draft of the paper and send it to me in early January. Thanks, Eben From: Yushuf Sharker <notifications@github.com> Reply-To: ekenah/transtat <reply@reply.github.com> Date: Thursday, December 21, 2017 at 11:51 AM To: ekenah/transtat <transtat@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: [ekenah/transtat] LR Confidence interval for unobs_tr despite vcov matrix is PD (#8) data.zip<https://github.com/ekenah/transtat/files/1579862/data.zip> library(transtat) pdata <- read.csv("pdata.csv") xdata <- read.csv("xdata.csv") coef <- read.csv("coef.csv", header = TRUE) pairwise data to be combined with external data pdat <- pdata pdat$interval <- pdat$exit - pdat$entry pdat$external <- 0 pdat$trace <- 1 pdat$start <- 0 pdat$stop <- pdat$interval external data xdat <- xdata names(xdat)[names(xdat) == "x"] <- "susx" xdat$external <- 1 xdat$infectious <- NA xdat$infx <- 0 xdat$interval <- xdat$exit # entry is at time 0 xdat$start <- xdat$entry xdat$stop <- xdat$exit dat <- rbind(pdat, xdat) #Transreg works unobs_tr <- transreg( Surv(start, stop, infset) ~ infx + susx + ext(external), sus = "susceptible", data = dat, subset = (trace == 1), dist = "weibull", xdist = "exponential" ) #Variance-covariance matrix is PD eigen(vcov(unobs_tr))$values >.00 #confint keep working for long time, I have waited overnight. ci_unobs_tr_lr <-confint(unobs_tr, type = 'lr') #This errors are not informative confint(unobs_tr, "log(shape)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed confint(unobs_tr, "(Intercept)", type = 'lr') Error in if (!is.numeric(lower) || !is.numeric(upper) || lower >= upper) stop("lower < upper is not fulfilled") : missing value where TRUE/FALSE needed — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#8>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mZKgKG1Ys0BB8YU8Xx80joWSeguBks5tCoyqgaJpZM4RKFLz>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ekenah_transtat_issues_8-23issuecomment-2D353400924&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=BVF4DRIGRhK7DP1u7K_48nxzPxZPSdRewh2pQX1EvM0&e=>, or mute the thread<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEkA32lmsmeFZslLXYS5IYUeWS0c2Gvuks5tCo0JgaJpZM4RKFLz&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=3P-w-te4SPHWJRwJAiEcTiCHm8LtfrGZVtsJUYOYxhQ&s=eCg8am9Touy2XLkS7Hx1E6YXIfEAsWcliCFOgShKWaM&e=>. — You are receiving this because you commented. Reply to this email directly, view it on GitHub<#8 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mXCA3UCtuhzrRwYixZetg_KtHE2Xks5tCo5ggaJpZM4RKFLz>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ekenah_transtat_issues_8-23issuecomment-2D353447123&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=Cs4-oDYX8ujQoVeGnHBFeGvH5IBcuklbJP0DS3RZbGs&s=alw5NcEMCDRNAkEeMi241R-_AOBNbmXcMPbyewpBbtM&e=>, or mute the thread<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AEkA36ounvnkkEd9BvK0tsHqXX24jANUks5tCrwQgaJpZM4RKFLz&d=DwMFaQ&c=cjytLXgP8ixuoHflwc-poQ&r=JcpGfI-4UWGF3Jb9A__sPJ2O1mFzDH_XkInJTMmZ9G0&m=Cs4-oDYX8ujQoVeGnHBFeGvH5IBcuklbJP0DS3RZbGs&s=VSDADrm0EDRSK63YNyrc3XXLFBbiqDNFzatHIaXen3A&e=>. — You are receiving this because you commented. Reply to this email directly, view it on GitHub<#8 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFP4mShnc0imb2V8Qq8SRks-HQQfyE5Fks5tC7wKgaJpZM4RKFLz>.