sum_hills with a multi CV COLVAR file (bug?)

[khala...@gmail.com]

Hello, everyone.
I'm trying to build a 2D free energy diagram out of a COLVAR file that has records for 7 CVs.
The default options work, but are very slow. My sigmas are tiny, so sum_hills guesses lots of bins.
I'd like to set the number of bins/min/max/spacing manually, but the code seems to expect different
numbers of arguments for those settings at different times during execution.

$plumed sum_hills --histo COLVAR --mintozero --outhisto fes_2D.dat --kt 2.4943500 --sigma 0.02,0.01 --idw cv1,cv2 --min 0,0 --max 10,3 --spacing 0.1,0.02

+++ Internal PLUMED error
+++ file core/../../core/CLTool.h, line 140
+++ message: vector read in for keyword  --min has wrong size

I put a printf statement right before to check what condition was causing this and turns out the input
parser expects bins/min/max/spacing to have 7 arguments, one for each CV in the file. So I tried:

$plumed sum_hills --histo COLVAR --mintozero --outhisto fes_2D.dat --kt 2.4943500 --sigma 0.02,0.01 --idw cv1,cv2 --min 0,0,0,0,0,0,0 --max 10,1,1,1,1,1,1 --spacing 0.1,1,1,1,1,1,1

+++ Internal PLUMED error
+++ file Grid.cpp, line 45
+++ message: assertion failed args.size()==gmin.size(), grid min dimensions in input do not match number of arguments

Similar things happen when trying to do a 1D energy surface.

$plumed sum_hills --histo COLVAR --mintozero --outhisto fes.dat --kt 2.4943500 --sigma 0.02 --idw cv1 --spacing 0.1

+++ Internal PLUMED error
+++ file core/../../core/CLTool.h, line 140
+++ message: vector read in for keyword  --spacing has wrong size

$plumed sum_hills --histo COLVAR --mintozero --outhisto fes.dat --kt 2.4943500 --sigma 0.02 --idw cv1 --spacing 0.1,1,1,1,1,1,1

...
PLUMED: Action FUNCSUMHILLS
PLUMED:   with label @7
PLUMED:   with arguments cv1 cv2 cv3 cv4 cv5 cv6 cv7
PLUMED:   Output format is %14.9f
PLUMED:   The number of bins will be estimated from GRID_SPACING
[*] *** Process received signal ***
[*] Signal: Segmentation fault (11)
[*] Signal code: Address not mapped (1)
[*] Failing at address: 0x8
...

Default binning settings work though.

I'm not sure if this is a bug or I'm doing something wrong.
I've seen this behavior with tags v2.3.1, v.2.4.0, and v2.4.1 cloned from github, so it's not something new.
Minimal COLVAR file to reproduce the problem is attached.

Cheers,
Yuriy Khalak.

[Giovanni Bussi]

Hi,

how many CVs are you biasing with metadynamics?

In case you are just trying to make a histogram, notice that you could just edit the COLVAR file in order to have only the variables that you need.

In case METAD is biasing 7 CVs, then I would say they are too many

Giovanni

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users+unsubscribe@googlegroups.com.
To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/4e9deede-bead-4d76-ab90-9278e8b38d10%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[khala...@gmail.com]

Hi, Giovanni, and thanks for answering.
It's from an unbiased replicate in BEMD. I'll reconstruct the file with just the relevant CVs. You are right, it's probably the easiest way to deal with this.

Thanks,
Yuriy.

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

[Billy Noonan]

Hi There,

Have noticed that this bug still exists 5 years later. Thank you for the solution! :) 

Kind Regards,

Billy