Matlab configuration advice - Revision history

Kuprov: Normalise sequential blank lines

2026-04-25T11:27:13Z

Normalise sequential blank lines

2018-07-31T12:08:32Z

‎Parallelization

2018-01-01T15:52:07Z

2017-01-09T15:05:28Z

‎Batch runs on PBS clusters

2016-12-04T11:06:33Z

‎Batch runs on PBS clusters

2016-12-04T11:05:19Z

‎Batch runs on PBS clusters

2016-12-04T11:03:03Z

‎Batch runs on PBS clusters

2016-11-20T16:12:30Z

‎Batch runs on PBS clusters

2016-11-20T16:07:32Z

‎Batch runs on PBS clusters

2016-11-17T18:08:39Z

‎Batch runs on PBS clusters

← Older revision		Revision as of 11:27, 25 April 2026
Line 70:		Line 70:

	You can get your current path by running the ''path'' command in Matlab. Do not forget to load the Matlab module in PBS before you submit the script to the PBS queue. The module should be installed by your technical support team.		You can get your current path by running the ''path'' command in Matlab. Do not forget to load the Matlab module in PBS before you submit the script to the PBS queue. The module should be installed by your technical support team.
−

	''Version 2.1, authors: [[Ilya Kuprov]]''		''Version 2.1, authors: [[Ilya Kuprov]]''

@@ Line 3: / Line 3: @@
 ==Parallelization==
-If you plan to run lengthy parallel calculations, disable the parallel pool timeout (Matlab Ribbon / Parallel / Parallel Preferences) by unticking the "shut down and delete" box in the settings.
+If you plan to run lengthy parallel calculations, disable the parallel pool timeout (Matlab Ribbon / Parallel / Parallel Preferences) by unticking the "shut down and delete" box in the settings. If this is not done, Matlab sometimes kills the parallel pool in the middle of a calculation.
 ==Java heap size==

← Older revision		Revision as of 15:52, 1 January 2018
Line 71:		Line 71:
	You can get your current path by running the ''path'' command in Matlab. Do not forget to load the Matlab module in PBS before you submit the script to the PBS queue. The module should be installed by your technical support team.		You can get your current path by running the ''path'' command in Matlab. Do not forget to load the Matlab module in PBS before you submit the script to the PBS queue. The module should be installed by your technical support team.

−	''~~Revision 3284~~, authors: [[Ilya Kuprov]]''	+
		+	''Version 2.1, authors: [[Ilya Kuprov]]''

@@ Line 37: / Line 37: @@
       cd $PBS_O_WORKDIR
-      matlab -nodesktop -r "name_of_your_function(`expr ${PBS_ARRAYID}`); exit" 1>output`expr ${PBS_ARRAYID}`.log 2>/dev/null </dev/null
+      matlab -nodesktop -r "your_function(`expr ${PBS_ARRAYID}`); exit"   \
-This script assumes that you have a Matlab function called name_of_your_function.m in your working directory, that this function takes a scalar integer from 1 to 22 as a job identifier and that you would need a 16-core node for 59 hours for each of those jobs. The header of name_of_your_function.m should contain a manual call to start the local parallel pool:
+This script assumes that you have a Matlab function called your_function.m in your working directory, that this function takes a scalar integer from 1 to 22 as a job identifier and that you would need a 16-core node for 59 hours for each of those jobs. The header of your_function.m should contain a manual call to start the local parallel pool, for example:
       parpool('local',16);

@@ Line 29: / Line 29: @@
       #!/bin/bash
       #PBS -l nodes=1:ppn=16
-      #PBS -l walltime=119:00:00
+      #PBS -l walltime=59:00:00
       #PBS -V
       #PBS -m n
@@ Line 39: / Line 39: @@
       matlab -nodesktop -r "name_of_your_function(`expr ${PBS_ARRAYID}`); exit" 1>output`expr ${PBS_ARRAYID}`.log 2>/dev/null </dev/null
-This script assumes that you have a Matlab function called name_of_your_function.m in your working directory, that this function takes a scalar integer from 1 to 22 as a job identifier and that you would need a 16-core node for 119 hours for each of those jobs. The header of name_of_your_function.m should contain a manual call to start the local parallel pool:
+This script assumes that you have a Matlab function called name_of_your_function.m in your working directory, that this function takes a scalar integer from 1 to 22 as a job identifier and that you would need a 16-core node for 59 hours for each of those jobs. The header of name_of_your_function.m should contain a manual call to start the local parallel pool:
       parpool('local',16);

@@ Line 66: / Line 66: @@
               '~/spinach/kernel/pulses');
-You can get your current path by running the ''path'' command in Matlab.
+You can get your current path by running the ''path'' command in Matlab. Do not forget to load the Matlab module in PBS before you submit the script to the PBS queue. The module should be installed by your technical support team.
 ''Revision 3284, authors: [[Ilya Kuprov]]''

@@ Line 64: / Line 64: @@
               '~/spinach/kernel/optimcon',...
               '~/spinach/kernel/overloads',...
-              '~/spinach/kernel/pulses',...
+              '~/spinach/kernel/pulses');
 You can get your current path by running the ''path'' command in Matlab.
 ''Revision 3284, authors: [[Ilya Kuprov]]''