need your opinion: cancer center scenario

[mokoka...@gmail.com]

Dear Experts, need your opinion..

A center for treatment of leukemia has only 30 beds

Treatment involves the following process:

Induction 1--> Induction 2 --> Consolidation 1 --> Consolidation 2

Step #1 --> Induction 1

-accepting new cases to receive Chemotherapy Known as Induction 1 Regimen

-rate of arrival 2.3 patients / week

-average length of stay (ALOS) = 31 days

Step #2 -->  Induction 2

-84.2 % of patients from step #1 arrive successfully to this step

-ALOS = 28 days

-highest priority for admission

Step #3  --> Consolidation 1

-74.6% of patients from step#1 arrive successfully to this step

-ALOS = 25 days

- priority for admission less than step#2 but higher than step#1

Step #4 --> Consolidation 2

-73.5% of patients from step#1 arrive successfully to this step

-ALOS = 31.5 days

- priority for admission less than step#2 but higher than step#1

 

Using the R-script below, running the simulation for 2 years (730 days) and 30 beds gave the following

server status: 30(30) | queue status: 42

> mean(x$queue)

[1] 22.3562

 

Using the R-script below, running the simulation for 2 years (730 days) and 31 beds gave the following

server status: 31(31) | queue status: 13

> mean(x$queue)

[1] 6.314767

Using the R-script below, running the simulation for 2 years (730 days) and 32 beds gave the following

server status: 30(32) | queue status: 0

> mean(x$queue)

[1] 0.57

Does this mean that: "if I increase the center capacity from 30 to 32 beds, I will have no waiting list?"

========================= R Script ==============================

 

library(simmer)

# rate of new patient arrival = 2.3 per week

# inter-arrival time (IAT)= 7/2.3 days

#rn = interarrival time of new cases who will receive Induction 1

rn <- 7/2.3 # IAT = 7/2.3, Priority = 1

#rind2 = interarrival time of patients who will receive Induction 2

rind2 <- (1/0.842) * rn # priority = 9

#consol1 = interarrival time of patients who will receive consolidation 1

rconsol1 <- (1/0.746) * rn # priority = 8

#consol1 = interarrival time of patients who will receive consolidation 2

rconsol2 <- (1/0.735) * rn # priority = 8

env <- simmer("AML-env")

## add a new patient trajectory

newpatient <- create_trajectory("newAMLpatientpath") %>%

  seize("bed", amount = 1, priority = 1) %>%

  ## 31 day

  timeout(function() rnorm(1, 31, 2)) %>% 

  release("bed", 1) 

## add a Induction2 patient trajectory

rind2patient <- create_trajectory("rind2AMLpatientpath") %>%

  seize("bed", amount = 1, priority = 9) %>%

  ## 28 days

  timeout(function() rnorm(1, 28, 2)) %>%

  release("bed", 1)

## add a Consolidation1 patient trajectory

consol1patient <- create_trajectory("consol1AMLpatientpath") %>%

    seize("bed", amount = 1, priority = 8) %>%

  ## 25 days

  timeout(function() rnorm(1, 25, 2)) %>%

  release("bed", 1)

## add a Consolidation2 patient trajectory

consol2patient <- create_trajectory("consol2AMLpatientpath") %>%

    seize("bed", amount = 1, priority = 8) %>%

  ## 31.5 days

  timeout(function() rnorm(1, 31.5, 2)) %>%

  release("bed", 1)

env %>%

  add_resource("bed", 30) %>%

  add_generator("newpatient", newpatient, function() rn) %>%

  add_generator("rind2patient", rind2patient, function() rind2) %>%

  add_generator("consol1patient", consol1patient, function() rconsol1) %>%

  add_generator("consol2patient", consol2patient, function() rconsol2) 

env %>% run(until=730) ## run for two years (365 * 2 = 730)

x<-env %>% get_mon_resources()

y<-env %>% get_mon_arrivals()

mean(x$queue)

[Iñaki Úcar]

Does this mean that: "if I increase the center capacity from 30 to 32 beds, I will have no waiting list?"

It would, if the model were good, but it doesn't seem reasonable to me.

• You have no arrival process. You've merely assigned constant IATs.
• Patients from the first stage don't go through subsequent stages. Instead, you are generating new different arrivals for each stage (!).

[Mohamed Kamal]

Thanks a lot for your response. I did not know it is possible to make patients go through subsequent stages in r-simmer. How is this possible ? In the examples provided with documentation I could not find an example of entities in sequence. I might need your help, if possible.

Regarding IATs I can use a distribution (prefer exponential ?)

[Iñaki Úcar]

Thanks a lot for your response. I did not know it is possible to make patients go through subsequent stages in r-simmer. How is this possible ? In the examples provided with documentation I could not find an example of entities in sequence. I might need your help, if possible.

Building a trajectory with all the steps needed. The introductory vignette has an example in which patients visit a nurse, then a doctor and then the administration.

 

Regarding IATs I can use a distribution (prefer exponential ?)

You need to fit a distribution based on real data. Maybe the exponential is a good candidate, maybe not.

 

--

Iñaki Úcar
http://www.enchufa2.es
@Enchufa2

[Mohamed Kamal]

The problem is that I have only one Resource = "Bed" I have 30 beds to accommodate all patients in all stages 

[Mohamed Kamal]

To clarify more, in real-life I have a long waiting list of patients waiting for a "free" bed. Patients from new cases, Induction2, consolidation1 and consolidation2 are all competing for the same and only resource (30 Bed). So, if I design a trajectory, I will not be able to "distribute" beds over different stages. 

[Bart Smeets]

You could built one trajectory to describe your simulation problem. Have look at the branching tactics described in https://cran.r-project.org/web/packages/simmer/vignettes/C-trajectories.html#branch. From a branch statement you can then model when and how you move towards a subtrajectory (e.g. the "74.6% of patients from step#1 arrive successfully to this step" statements).

Between each step you can use a timeout where you wait an appropriate amount of time between e.g. step 1 and step 2. Once the timeout is over, you can let the patient re-seize a bed (and it will compete here with other patients that might be in a different step). You can of course also work with priorities in the seize, to give different patients at different stages different priorities at seizing a bed (which is probably realistic)

Op woensdag 8 juni 2016 14:51:31 UTC+2 schreef Mohamed Kamal:

[Mohamed Kamal]

thanks Bart and Iñaki

I will rewrite it and post it back on the group

--
You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/93bebb04-1325-4f1d-9a91-292b352cc38a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP

[Mohamed Kamal]

please check the following modified script with branches, your feedback is highly appreciated...

library(simmer)

# rate of new patient arrival = 2.3 per week

# interarrival time (IAT)= 7/2.3 days

env <- simmer("AML-env")

## add a new patient trajectory

newpatient <- create_trajectory("newAMLpatientpath") %>%

  branch(function() sample(c(1,2),size=1,replace=F,prob=c(0.999,0.001)), merge=c(T, T), 

         create_trajectory("Induction1") %>%

           seize("bed", amount = 1, priority = 1) %>%

           seize("Induction1_in", amount = 1, priority = 1) %>%

           ## 31 day

           timeout(function() rnorm(1, 31, 2)) %>% 

           release("bed", 1)%>%

           

           #================ Induction 2

           branch(function() sample(c(1,2),size=1,replace=F,prob=c(0.842,0.158)), merge=c(T, T), 

                  create_trajectory("Induction2") %>%

                    seize("bed", amount = 1, priority = 9) %>%

                    seize("Induction2_in", amount = 1, priority = 9) %>%

                    ## 28 day

                    timeout(function() rnorm(1, 28, 2)) %>% 

                    release("bed", 1)%>%

                    #=============== Consolidation 1

                    branch(function() sample(c(1,2),size=1,replace=F,prob=c(0.746,0.254)), merge=c(T, T), 

                           create_trajectory("Consolidation1") %>%

                             seize("bed", amount = 1, priority = 8) %>%

                             seize("Consolidation1_in", amount = 1, priority = 8) %>%

                             ## 25 day

                             timeout(function() rnorm(1, 25, 2)) %>% 

                             release("bed", 1) %>%

                             #===============Consolidation 2

                             branch(function() sample(c(1,2),size=1,replace=F,prob=c(0.99,0.01)), merge=c(T, T), 

                                    create_trajectory("Consolidation2") %>%

                                      seize("bed", amount = 1, priority = 8) %>%

                                      seize("Consolidation2_in", amount = 1, priority = 8) %>%

                                      ## 31.5 day

                                      timeout(function() rnorm(1, 31.5, 2)) %>% 

                                      release("bed", 1)

                                    ,

                                    create_trajectory("Consolidation2_out") %>%

                                      seize("Consolidation2_out", amount = 1, priority = 1) 

                                    

                             ) 

                           #=============End Consolidation2

                           ,

                           create_trajectory("Consolidation1_out") %>%

                             seize("Consolidation1_out", amount = 1, priority = 1) 

                           

                    ) 

                  #===============End Consolidation 1

                  ,

                  create_trajectory("Induction2_out") %>%

                    seize("Induction2_out", amount = 1, priority = 1) 

                  

           ) 

         #=================End Induction 2

         

         ,

         create_trajectory("Induction1_out") %>%

           seize("Induction1_out", amount = 1, priority = 1) 

         

  ) 

env %>%

  

  ## add true resource

  add_resource("bed", 20) %>%

  

  ## add resources for counting purpose

  add_resource("Induction1_in", Inf) %>%

  add_resource("Induction2_in", Inf) %>%

  add_resource("Consolidation1_in", Inf) %>%

  add_resource("Consolidation2_in", Inf) %>%

  

  add_resource("Induction1_out", Inf) %>%

  add_resource("Induction2_out", Inf) %>%

  add_resource("Consolidation1_out", Inf) %>%

  add_resource("Consolidation2_out", Inf) %>%

  

  

  add_generator("newpatient", newpatient, function() rnorm(1, mean=3.33, sd=0.5)) 

env %>% run(until=730) ## run for two years (365 * 2 = 730)

x<-env %>% get_mon_resources()

y<-env %>% get_mon_arrivals()

z<-env %>% get_mon_attributes()

plot_resource_utilization(env, c("bed"))

write.csv(y,"y.csv")

mean(x$queue)

arrivals <- get_mon_arrivals(env, per_resource = T)

write.csv(arrivals,"arrivals.csv")

[Mohamed Kamal]

I also need to know if there is a way to breakdown the queue. 

thanks again in advance.

--

You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/99e9da8e-bbb8-4333-8d92-ff240091dae4%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Iñaki Úcar]

What do you mean by "breakdown the queue"?

You received this message because you are subscribed to the Google Groups "simmer-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0SdiONjj03iFckO0yAuqmV02SoWsgxtv3dLxyd%2Bq0KcYQ%40mail.gmail.com.

[Mohamed Kamal]

Dear Iñaki

thanks a lot for your efforts and support

regarding queue, I mean percentage of queue from each stage: how many patients are waiting for Induction1, how many for Induction2, how many for Consolidation1 and consolidation2. All what I get is total number in queue.

one more thing, is my model OK now or do you still see something that needs improvement?

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq3Z2HZ__Z%2Bgg3CKzVkdymXaSYMmzxP8yzcQHH4Wyu%3DYUQ%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Iñaki Úcar]

Sorry, I can't take a look at your code until tomorrow.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0Q%2BOSPwk4tchrxGcMoN_Ypu8V-iHDWccgdOZRLyt1Epfw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

[Mohamed Kamal]

@Bart Smeet

any feedback

To unsubscribe from this group and all its topics, send an email to simmer-devel+unsubscribe@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/99e9da8e-bbb8-4333-8d92-ff240091dae4%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP

--
You received this message because you are subscribed to the Google Groups "simmer-devel" group.

To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel+unsubscribe@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0SdiONjj03iFckO0yAuqmV02SoWsgxtv3dLxyd%2Bq0KcYQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.

To unsubscribe from this group and all its topics, send an email to simmer-devel+unsubscribe@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq3Z2HZ__Z%2Bgg3CKzVkdymXaSYMmzxP8yzcQHH4Wyu%3DYUQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP

--
You received this message because you are subscribed to the Google Groups "simmer-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0Q%2BOSPwk4tchrxGcMoN_Ypu8V-iHDWccgdOZRLyt1Epfw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Mohamed Kamal]

Experts

need your advice, is it possible to label Beds as Bed01, Bed02 ,....., Bed30

and when the patient try to seize "loop" through the beds and choose the first free bed

[Bart Smeets]

Haven't looked at it yet, will take a look if I find some free time. Is there anything in the output that makes you think its "incorrect"?

Op donderdag 9 juni 2016 11:56:00 UTC+2 schreef Mohamed Kamal:

[Bart Smeets]

I think this related to https://github.com/r-simmer/simmer/issues/52

Op donderdag 9 juni 2016 15:39:36 UTC+2 schreef Mohamed Kamal:

[Mohamed Kamal]

the output is OK, after running the simulation it is almost same as real data
Waiting for your feedback


but I need (optionally) to label beds with numbers, is it possible ?

>>>>>>> simmer-devel...@googlegroups.com.
>>>>>>> To post to this group, send email to simmer...@googlegroups.com.
>>>>>>> To view this discussion on the web visit

>>>>>>> https://groups.google.com/d/msgid/simmer-devel/99e9da8e-bbb8-4333-8d92-ff240091dae4%40googlegroups.com.
>>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Mohamed Kamal, R.Ph., CHRP
>>>>>>
>>>>>> --
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "simmer-devel" group.
>>>>>> To unsubscribe from this group and stop receiving emails from it, send

>>>>>> an email to simmer-devel...@googlegroups.com.
>>>>>> To post to this group, send email to simmer...@googlegroups.com.
>>>>>> To view this discussion on the web visit

>>>>>> https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0SdiONjj03iFckO0yAuqmV02SoWsgxtv3dLxyd%2Bq0KcYQ%40mail.gmail.com.
>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>> --
>>>>> You received this message because you are subscribed to a topic in the
>>>>> Google Groups "simmer-devel" group.
>>>>> To unsubscribe from this topic, visit
>>>>> https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
>>>>> To unsubscribe from this group and all its topics, send an email to

>>>>> simmer-devel...@googlegroups.com.
>>>>> To post to this group, send email to simmer...@googlegroups.com.
>>>>> To view this discussion on the web visit
>>>>> https://groups.google.com/d/msgid/simmer-devel/CALEXWq3Z2HZ__Z%2Bgg3CKzVkdymXaSYMmzxP8yzcQHH4Wyu%3DYUQ%40mail.gmail.com.
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>>
>>>>
>>>> --
>>>> Mohamed Kamal, R.Ph., CHRP
>>>>
>>>> --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "simmer-devel" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to simmer-devel...@googlegroups.com.
>>>> To post to this group, send email to simmer...@googlegroups.com.
>>>> To view this discussion on the web visit
>>>> https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0Q%2BOSPwk4tchrxGcMoN_Ypu8V-iHDWccgdOZRLyt1Epfw%40mail.gmail.com.
>>>>
>>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>>
>>>
>>>
>>> --
>>> Iñaki Úcar
>>> http://www.enchufa2.es
>>> @Enchufa2
>

> --
> You received this message because you are subscribed to a topic in the
> Google Groups "simmer-devel" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to

> simmer-devel...@googlegroups.com.
> To post to this group, send email to simmer...@googlegroups.com.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/simmer-devel/7db4c8b0-3bdb-4896-bed8-0f0024474cd0%40googlegroups.com.

[Mohamed Kamal]

by the way, please ignore the very first percentages i sent in the very first post. the latest numbers are more accurate

[Mohamed Kamal]

any feedback ?

[Bart Smeets]

Can you explain why you want to label the beds in this case? If the first free bed will always be seized, won't the final results be the comparable to the output you have now?

Op donderdag 9 juni 2016 15:39:36 UTC+2 schreef Mohamed Kamal:

Experts

[Mohamed Kamal]

I think I dont "need" to label beds, just asking if it is possible to do it.

So, do you think this model is OK ? I mean, if the percentages are matching reality, shall i depend on it to predict scenarios?

To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/99e9da8e-bbb8-4333-8d92-ff240091dae4%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP

--
You received this message because you are subscribed to the Google Groups "simmer-devel" group.

To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0SdiONjj03iFckO0yAuqmV02SoWsgxtv3dLxyd%2Bq0KcYQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.

To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq3Z2HZ__Z%2Bgg3CKzVkdymXaSYMmzxP8yzcQHH4Wyu%3DYUQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP

--
You received this message because you are subscribed to the Google Groups "simmer-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0Q%2BOSPwk4tchrxGcMoN_Ypu8V-iHDWccgdOZRLyt1Epfw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Iñaki Úcar
http://www.enchufa2.es
@Enchufa2

--

You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.

To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/e45765c7-1782-4afb-ae68-b3b7e2d3349b%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Shawn Garbett]

On Fri, Jun 10, 2016 at 7:08 AM, Mohamed Kamal <mohamedk...@gmail.com> wrote:

| So, do you think this model is OK ? I mean, if the percentages are matching reality, shall i depend on it to predict scenarios?

What do you mean matching reality? Model validation is important. Did you do a statistical prop test or ANOVA? If so, what were the results, etc? How much data do you have from reality?

Then if the model is validated for known parameters, to change those to areas outside observed usually results in extrapolation. It's important to be aware of the dangers of extrapolation. The changes made, may change external behaviors, or there may be factors that come into play that weren't part of the original model. With this cautionary note, it's still a powerful technique. Make sure that the final statistical method used is for computing predicted intervals and not confidence intervals.

---

Shawn Garbett

Vanderbilt Biostatistics

[Iñaki Úcar]

To avoid a branch inside a branch inside a branch… I would simulate something like this:

patient <- create_trajectory() %>%
  seize("waiting_induction1", 1) %>%
  seize("bed", 1, priority=induction1) %>%
  release("waiting_induction1", 1) %>%
  timeout(induction1_timeout) %>%
  branch(induction1_prob_continue, c(T, F),
         create_trajectory() %>% release("bed"),
         create_trajectory() %>% release("bed")) %>%
  seize("waiting_induction2", 1) %>%
  seize("bed", 1, priority=induction2) %>%
  release("waiting_induction2", 1) %>%
  timeout(induction2_timeout) %>%
  ...

and so on, where “waiting_induction1”, “waiting_induction2”… are resources with infinite capacity allocated to count how many patients are waiting for each stage.

​

--

You received this message because you are subscribed to the Google Groups "simmer-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAJ2peNeD00hUgjXJDfWFkeRoqazGwKFKjkqB9iQDBVYmuM-X0g%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Thanks a lot. Would you kindly advise me how to get waiting time data as vector. I can only get it as curve using function plot_evolution_arrival_times(env, "waiting_time")

You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq2ZcXT6kOtA07_Fz1r8hN1QUX6B_cSY04C5KROaEsPoDw%40mail.gmail.com.

[moko kamello]

Is it possible to use the following to get the waiting time:

arrivals_no_resource<-get_mon_arrivals(env)

arrivals_no_resource$waiting<-arrivals_no_resource$end_time-arrivals_no_resource$start_time-arrivals_no_resource$activity_time

[moko kamello]

Dear Experts, please allow me to ask one more question

what if the patient is discharged from the hospital for 5 days before re-admission. I don't want to count those 5 days as waiting time (because it is not). How is it possible to do it in simmer?

[Iñaki Úcar]

Yes, that's the waiting time in general: end_time - start_time - activity_time.

But specifically in my example, the infinite resources waiting_induction1, waiting_induction2, etc., measure just your waiting time in queue. So the activity_time column from these resources is your waiting time.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cnPPR%2BhGarRzQsffGws7C_tbbj0%3DZ-C6g%3D9T8S8%2BkKyVA%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Iñaki Úcar]

what if the patient is discharged from the hospital for 5 days before re-admission. I don't want to count those 5 days as waiting time (because it is not). How is it possible to do it in simmer?

I don't know if I follow you. You mean between two stages? Then you can simply set a timeout for 5 days between stages, between the release from one stage and the seize from the next stage.

[mokoka...@gmail.com]

Thanks a lot. 

[mokoka...@gmail.com]

if i need to generate 10 patients at once, should i put the add_generator function in a loop

for (i in 1:10){

add_generator("newpatient", newpatient, function() rexp(1, 1/myIAT))

}

OR there is another solution?

[Iñaki Úcar]

No, you shouldn’t. You need, instead of function() rexp(1, 1/myIAT), a new function returning one time rexp(1, 1/myIAT) and 9 times 0, one time rexp(1, 1/myIAT), 9 times 0, and so on.

​

--

You received this message because you are subscribed to the Google Groups "simmer-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/1fff68d7-aea2-46ff-b15c-82d200788ac8%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Mohamed Kamal]

hmmm, is it possible (when you have time) to show me an example.

You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq1Gnb6y1jnni0YT8nqaUnx96Ot7JA9QeUwAQ1E8EC9Wvg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP

[Shawn Garbett]

Brute force using a global (there's certainly a better way):

tenatatime_cnt <- 0

tenatatime <- function()

{

  if(tenatatime_cnt == 0)

  {

    tenatatime_cnt <<- 9

    rexp(1)

  }

  else

  {

    tenatatime_cnt <<- tenatatime_cnt - 1

    0

  }

}

env %>% add_generator("patient", traj, tenatatime)

Note this goes on forever now. 

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0Q78Zpd7nfJcPp56yAKTnn5YX6v8VKAPu5C-7V4kCp%3Dgw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Shawn Garbett
615.397.8737

[Iñaki Úcar]

Thanks Shawn. That does the trick.

However, this is easier with the current version (check out the master branch on GitHub), because the generation function can return more than one value at once (see #65), so now we can do the following:

add_generator("newpatient", newpatient, function() c(rexp(1, 1/myIAT), rep(0, 9)))

​

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CAJ2peNfj2m5nmpHykYmRfD0d03XaYj%2B3XBFDzZzBcBVkZZFHUg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

I don't know if it is my bad, but i get the following

add_generator("newpatient", newpatient, function() c(rexp(1, 1/myIAT),
rep(0, 9)))

Error: expecting a single value

Thanks Shawn and Iñaki

> https://groups.google.com/d/msgid/simmer-devel/CALEXWq0696it7s79KGtPzQqu-y8pYCwnQ6Xxzx8EGyrFkhR8mQ%40mail.gmail.com.

[Iñaki Úcar]

I said "with the current version on GitHub". You have to install it.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cmjtnzXq%2B7WBXU9VfzZmo0GH-DU3HEX6oHjzX_MSrs3LQ%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Mohamed Kamal]

what about if I need to simulate spread of infection ? Is there a way that a patient "infect" another patient (or more ). 

one idea is to ceate a resource "infected" with unlimited capacity for counting purpose, but his will not simulate the interaction of patient to patient.

any other ideas?

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq1LgzFBkj5vgjjb_2DtUHg8mg%2BE0OtGKfCGOSHfD8GnbQ%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Is it possible that "waiting time" be negative values ?

> https://groups.google.com/d/msgid/simmer-devel/CAGJu%2B0SB8CuOmJUTmhKR_G%2B8qHt7GgZckyMfin0Ai_EB0Y0Hzg%40mail.gmail.com.

[Iñaki Úcar]

2016-06-18 1:05 GMT+02:00 moko kamello <mokoka...@gmail.com>:

Is it possible that "waiting time" be negative values ?

No, it's not.

Iñaki

[moko kamello]

I don't know if this is my bad: the waiting time is positive only if in case of get_mon_arrival(env, per_resources=F)

However, It gives me mixed positive and negative values in case of per_resource=T and I have absolutely no clue why this happens

Any Help ?

--

You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq05ta4V7zfr_FHDTjHVieH1wLUcrBb0QO%3DDJAKE%3DPL3BA%40mail.gmail.com.

[Iñaki Úcar]

It seems there's a bug in the development version. The version on CRAN gives correct results though. It would be great if you could reproduce this problem with a much simpler example. It would help me debug the issue.

--
You received this message because you are subscribed to the Google Groups "simmer-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simmer-devel...@googlegroups.com.

To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cm4qTj43iwoqB3ZPe3g4gBtL7om5vept%2BsqRGubJ7OWhw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Iñaki Úcar]

Actually the current CRAN version (3.2.1) does NOT give correct results either, sorry. It's the release 3.1.2 which gives the correct results.

[moko kamello]

Could you please send me a zip file to install it as I have no experience to install it from github

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq2R7kRftyPrd22sk29b5DYsq0Y3RY_S2hc3iHskAvX_dA%40mail.gmail.com.

[Iñaki Úcar]

You can download every release from the Releases page: https://github.com/r-simmer/simmer/releases

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47ckSzZ3XEczmrZkn3W7kaP8Vw0Vmcv7nYgXnfTyb4zPXsg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Thanks

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq2jFQ2E4FMjHugHCBAAb4Qh%2BJhKpnKgm32QS%3DmtdXJVcA%40mail.gmail.com.

[Iñaki Úcar]

I've opened an issue on GitHub. If anyone finds a smaller piece of code featuring this bug, please, post it there.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cnAGrUb4UyCVLK7qgaCZJ-%3D6XZ2GbAy%3D47008j_Jczq_g%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Is there any quick work around until this issue is resolved ?

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq1T0M5KGWFr6GmiwHbBEQ_ncuHS40eTpyPOZTxGd63f3A%40mail.gmail.com.

[Iñaki Úcar]

Use the version 3.1.2.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47ckGNSxUFONQzfyaTfrFcyPASi7dy%3D3-yrD4CK%2BzL3U0zg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Now, using the latest CRAN version, I think the error results from the cumulative sum of time within each branch, this will result in activity time to become greater and lead to -ve values. Just an opinion.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq2T_1N2d9eZQ%2BxWhAJupnE9wETOFuXPvUn4o8Bo_FkPXg%40mail.gmail.com.

[moko kamello]

Attached simpler example with cummulative time calculated for a branch

[moko kamello]

For a reason or another the only version working is the one on CRAN. Is there any snippet of script to prevent cumulative sum of timeout inside a branch

[Iñaki Úcar]

The bug is already fixed. Check out the master branch on GitHub. The CRAN version has the bug and does NOT work properly.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cmXewEc6ML_d%3D-ugvSqjOpPg7C0-smTK_6%3DzsdjusRhjg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[Iñaki Úcar]

BTW, in order to avoid so many nested branches in your model, you’ll be probably interested in the new activity called leave, which does exactly what you need: leave a trajectory with some probability.

​

[moko kamello]

Have any sample ?

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq1jqu7LVC38Vb7Bz-ycwXxk9tYis70oAk7hq2qkqJjZYA%40mail.gmail.com.

[Iñaki Úcar]

See the associated issue: https://github.com/r-simmer/simmer/issues/63

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cnsFs2yZ%3DFXBTDvV51vJzjyHt46BNDeGDAnu-oKLjDz0g%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

OK, I see. But I think that leave() cannot be used if I want the two branches to seize the resources at different priorities and timeouts for example ..

 branch(function() sample(c(1,2),size=1,replace=T,prob=c(0.2,0.8)), merge=c(T, T),
                                            create_trajectory("relapse") %>%
                                            seize("bed", amount = 1, priority = 1) %>%
                                            seize("relapse", amount = 1, priority = 1) %>%
                                            timeout(function() rnorm(1, 40, 2)) %>%
                                            release("bed", 1)
                                               ,

                                             create_trajectory("relapse_out") %>%
                                             seize("bed", amount = 1, priority = 9) %>%
                                             seize("relapse_out", amount = 1, priority = 1)  %>%
                                             timeout(function() rnorm(1, 10, 2)) %>%
                                             release("bed", 1)
                                             
                                      )

So, in such case there is no other option but to use branch inside a branch, right?

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq3XZ8SV3WSeS_fhU%3DNivBeSDYLONUWhzmZBNF%2B4HGd0-w%40mail.gmail.com.

[Iñaki Úcar]

I was talking about the points in your trajectory, at the end of each stage, where a patient may leave with some probability. The case you mention is different, of course.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cm-6%2B9qinhLM6ibwjjO5y8yGst3Shfwm06DqYDxbWFttA%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Congratulations for version 3.3 btw

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq0QZrH8oVp7hQ73bt5f6g8E-XwLvuRJ7GFZwchr_2eN8Q%40mail.gmail.com.

[moko kamello]

in get_mon_resources(env, per_resource=T) i see only finished=TRUE. Is
there a way to get the finished=FALSE?

[Iñaki Úcar]

• get_mon_resources does NOT have a per_resource argument nor a finished column in the resulting data frame.
• get_mon_arrivals does have a per_resource argument, but
  • per_resource=F (default) shows a finished column, and you’ll see TRUEs and FALSEs.
  • per_resource=T does NOT show a finished column.

​

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47c%3Dx8p2yWzCff_b_nHtr_iKH27L02mcxBqZF_4UG5cw%2BgA%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Thanks, but I always have "TRUE" only. If for example 200 patients generated I only have 160-170 records with finished= "true" but zero records with finished= false

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq2VSRRhMjg1nRdza36L8MQdN2pSKR2zv0RygRztJ%3DhK6w%40mail.gmail.com.

[Iñaki Úcar]

That's because no one got rejected.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47cm-Z5QFmpaATQyz4isJBrzdjyC4%2Bj1WJxgo%3Dxu0NrRfvg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

Sorry for bothering, how can a patient get rejected

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq3C8j65mfDNaf9joYa277PJCChFXwnY6a8-K6zu8X6Rug%40mail.gmail.com.

[Iñaki Úcar]

By trying to seize a resource with no space in queue or by using `leave` instead of `branch`. A `branch` is an alternative but perfectly valid trajectory.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALz47c%3DaPs2ZKtdjDqaahz4YXp-_Rxd3xx-Jz9S_FSXbt%3DnG_w%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

[moko kamello]

"waiting time evolution" plot is plotted from the waiting time
calculated from the get_mon_arrivals(per_resource=FALSE), right?

> https://groups.google.com/d/msgid/simmer-devel/CALEXWq0k_ce12uJ0tPtyuss%3DTyFpHGk0SefSvj%3D%3DEufHNX343w%40mail.gmail.com.

[Iñaki Úcar]

"waiting time evolution" plot is plotted from the waiting time
calculated from the get_mon_arrivals(per_resource=FALSE), right?

Yes, it is. You can check it by yourself: https://github.com/r-simmer/simmer/blob/master/R/plot.R#L117

[Mohamed Kamal]

thanks a lot

--

You received this message because you are subscribed to a topic in the Google Groups "simmer-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/simmer-devel/sXcFXvUDD1w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to simmer-devel...@googlegroups.com.
To post to this group, send email to simmer...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/simmer-devel/CALEXWq2ikTVO6cXVH-12Yiy43HcQFrs8g0q9EMC35fL_y_D4uw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Mohamed Kamal, R.Ph., CHRP