-
Notifications
You must be signed in to change notification settings - Fork 14
/
iv-advanced.Rmd
52 lines (43 loc) · 2.47 KB
/
iv-advanced.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
---
title: "04 Advanced `SpaDES` use"
author:
- "Alex M. Chubaty"
- "Eliot J. B. McIntire"
date: "`r strftime(Sys.Date(), '%B %d %Y')`"
output:
rmarkdown::html_vignette:
number_sections: yes
self_contained: yes
toc: yes
vignette: >
%\VignetteEngine{knitr::rmarkdown}
%\VignetteIndexEntry{04 Advanced `SpaDES` use}
%\VignetteDepends{SpaDES.core, SpaDES.tools}
%\VignetteKeyword{discrete event simulation, spatial simulation models}
%\VignetteEncoding{UTF-8}
bibliography: bibliography.bib
---
# Advanced `SpaDES` use
_This vignette is still a work in progress._
## Memory monitoring
While `profvis::profvis` is an essential tool for memory monitoring using deep R internals, it is often not sufficient for a discrete event situation.
For example, it may be useful to know the *peak memory use* of an event, as this may be the limiting step for setting up many parallel instances.
There is an experimental tool that gets triggered with `options("spades.memoryUseInterval" = xxx)` where `xxx` is a `numeric` in seconds, e.g., `0.2`. If this is set, and `future` and `future.callr` are installed, then whenever a `spades` call is made, the memory use will be assessed at that regular interval.
The procedure is:
1. spawn a future session (i.e., a parallel session) that runs `system('ps')` which lists all processes. It only keeps the process that represents the process ID of the main R session;
2. that `ps` call writes to a text file every `getOption('spades.memoryUseInterval')`;
3. if you ran this with a `spades` call, setting `options("spades.memoryUseInterval" = 0.5)` or some other interval (in seconds), it will read that text file into the `simList` at the end (`on.exit`) of the `spades` call (_doing this triggers a file deletion of the text file_);
4. the object is then in `sim$.memoryUse$obj`.
At that point, the function `memoryUse` can be called on the `simList` and it will do a join on the `sim$.memoryUse$obj` with the `completed(sim)` _by_ time stamp, so each event shows its memory use.
```{r memoryUse, eval=FALSE, echo=TRUE}
if (requireNamespace("future", quietly = TRUE) &&
requireNamespace("future.callr", quietly = TRUE)) {
options("spades.memoryUseInterval" = 0.5)
# run your simInit and spades calls here
# sim <- simInit()
# sim <- spades(sim)
memoryUse(sim, max = TRUE) # this should show peak memory use by eventType -- i.e., summarizes if multiple times
memoryUse(sim, max = FALSE) # this should show peak memory use by event
}
```
# References