Out of memory: 'Calloc' could not allocate memory (50000000 of 544 bytes) #243
Replies: 5 comments 6 replies
-
|
@namtien0312 , I am not certain if there is a way to reduce the memory usage for the integration steps within rxode2. The system does use many times more memory during the process than are output in the final data frame. All of the intermediate calculations take much more space than the final stored result. But, there are several work-arounds that you can consider:
|
Beta Was this translation helpful? Give feedback.
-
|
You can also try the development version of devtools::install_github("nlmixr2/rxode2") |
Beta Was this translation helpful? Give feedback.
-
|
I am transferring this to a discussion. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you two for the time to consider my case.
In my case, the target is AUC, so I need the necessary points to approximate the integral (as you described, 0.1) at a steady state (I have set ss = 1).
As you can see in my |
Beta Was this translation helpful? Give feedback.
-
|
If you want AUC, you can accomplish this by creating a state that integrates your concentration. Say you're dosing every 7 days and you're dosing 10 times. If Cc is the concentration of the central compartment (a secondary parameter) you can add another state called AUC. Then you can sample at the following times in days: The AUC calculated by the model is cumulative. So if you want the AUC over a specific interval you take the AUC at the end and subtract off the AUC at the beginning. So to calculate the AUC from day 21 to 28 you would do the following: AUC day 28 - AUC day 21. Then you only really need very few time points output. Just the ones at the beginning and the end of the interval where you want an AUC. |
Beta Was this translation helpful? Give feedback.
-
|
Ohh, thank you to suggest me a way to obtain AUC more swiftly and effectively. I have no idea that I can entirely get your point, especially with derivation. So if I have a model like that, I just add the term Thank again! |
Beta Was this translation helpful? Give feedback.
-
|
By the way, with antimicrobial, the target can be So in this case, whether there was a way to reduce the number of time points? |
Beta Was this translation helpful? Give feedback.
-
Yes, but the development version is not the one on CRAN. You need to download and compile from source. I am working toward putting it on CRAN soon, but it currently isn't there. |
Beta Was this translation helpful? Give feedback.
-
Well, you have to understand that the Here is an example (using library(rxode2)
#> Warning: package 'rxode2' was built under R version 4.2.1
#> rxode2 2.0.7 using 4 threads (see ?getRxThreads)
#> no cache: create with `rxCreateCache()`
one.compartment <- function() {
ini({
tka <- 0.45 # Log Ka
tcl <- 1 # Log Cl
tv <- 3.45 # Log V
eta.ka ~ 0.6
eta.cl ~ 0.3
eta.v ~ 0.1
add.sd <- 0.7
})
model({
ka <- exp(tka + eta.ka)
cl <- exp(tcl + eta.cl)
v <- exp(tv + eta.v)
d/dt(depot) = -ka * depot
d/dt(center) = ka * depot - cl / v * center
cp = center / v
d/dt(auc) <- cp
cp ~ add(add.sd)
})
}
e <- et(amt=100, ii=12, until=24) %>%
et(0,24, length.out=50)
s <- rxSolve(one.compartment, e)
#> ℹ parameter labels from comments will be replaced by 'label()'
plot(s, cp, auc)e2 <- e <- et(amt=100, ii=12, until=24, cmt="depot") %>%
et(amt=0, ii=12, until=24, cmt="auc", evid=5) %>% # replace the auc with zero
et(0,24, length.out=50)
s <- rxSolve(one.compartment, e2)
#> ℹ parameter labels from comments will be replaced by 'label()'
plot(s, cp, auc) |
Beta Was this translation helpful? Give feedback.
-
|
By the way, If I used My |
Beta Was this translation helpful? Give feedback.
-
|
Linux handles memory much differently than windows. If you have a large enough swap file, it should handle more simulations. |
Beta Was this translation helpful? Give feedback.
-
Dear Dr @mattfidler and
nlmixr/RxODEteam,When I used
nSub = 500, nStud = 50, cores = 4inrxSolvefunction, I got adataframeof194402216 bytesviaobject.size(my_df).When I used
nStud = 100, cores = 4instead, I got an error:And
error: cannot allocate vector of size 17.0 mbwhennStud = 55.When I incresed R memory by
memory.limit(size = 20000), it have not worked.My computer's RAM is 8 GB. How can I solve this issue by modifying some code rather than dealing with hardware? Thank you for your help!
My session info:
Beta Was this translation helpful? Give feedback.
All reactions