Skip to content

Commit faaa662

Browse files
committed
auto-commit
1 parent 17a39cb commit faaa662

File tree

10 files changed

+1526
-2
lines changed

10 files changed

+1526
-2
lines changed

.gitignore

Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,19 @@
1+
*.r~
2+
*~
3+
*autosave
4+
*swp
5+
*pdf
6+
*png
7+
*svg
8+
*rdata
9+
*RData
10+
*/#*
11+
.#*
12+
_*_.*
13+
*.dvi
14+
*.Rhistory
15+
*#
16+
*makefile0*
17+
*screenlog*
18+
* (copy)
19+
X.*

README.md

Lines changed: 33 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,36 @@
11
r-script-parallelisation
22
========================
33

4-
Template for splitting a job in an R script across multiple threads and computers
4+
Template for splitting a job in an R script across multiple threads and computers
5+
6+
7+
Context
8+
=======
9+
10+
One often wants to run a heavy R job across multiple threads and computers, like for doing simulations. Each simulation may require a set of parameters, and each simulation is carried out independently of each other. This template facilitates the splitting of the job, the execution of the job, and the harvest of results into a single file.
11+
12+
13+
Process
14+
=======
15+
16+
The data is in a single file, data.rdata, containing the rows id, and some parameters for each subjob. The threads and computers are specified in pars_parallel.r.
17+
18+
19+
Organisation
20+
============
21+
22+
data.rdata: initial data providing parameters for running each job
23+
run1.r: script to run action on slice. Requires to take as arguments a sequence of row ids (e.g. Rscript run1.r 1-10,21-30
24+
makefile: organises scripts and actions
25+
26+
27+
Requirements
28+
============
29+
30+
- R (>2.14)
31+
- GNU make
32+
- GIT
33+
- screen
34+
- ssh
35+
36+

0 commit comments

Comments
 (0)