Skip to content

AMDAT run accumulates memory despite delete commands #70

Description

@pkawak

I am really unfamiliar with these commands: delete_trajectory_list and delete_valuelist but I copied them over from an older script in the following workflow:

system_np
custom
./traj/trajectory_AGtestBinLJ8000_001_B5_1_S6_0_T334689.prd.custom

exponential 10 105 1.2 0 0 0.005
type1 6400 type2 1600
1 2
1 0
0 1

### ATOM SELECTION ###
#create a list of all atoms in the system
create_list all
all

#create a list of all atoms of type 1
create_list  type1_list
type_species type1 1

#create a list of all atoms of type 2
create_list  type2_list
type_species type2 2

### NEIGHBOR LIST CONSTRUCTION AND ANALYSIS ###
#create a neighbor list, for all particles, based upon the Voronoi tesselation.
create_voronoi_neighborlist vneighbors
list all

#create a neighbor list, for all particles, based on a distance cutoff of 1.4 times the interparticle contact distance matrix
create_distance_neighborlist dneighbors 1.4 ./inputs/sigma_matrix -1
list all

for frame 1 104

    #calculate the displacement magnitude of all particles at the time index of ${frame}, write out statistics on the distribution of these displacements, and save these displacements in a value list called displacement_list for later analysis.
    displacement_list data/neigh/displacement_stats_${frame}.dat displacement_list_${frame} ${frame}
    list all

    #output distribution statistics on the displacements computed above
    value_statistics data/neigh/displacement_dist_${frame}.dat displacement_list_${frame} 4

    #define a new trajectory list, called "fast_${frame}" by selecting those particles in the upper 6.5% of displacements stored in displacement_list.
    value_list threshold_percentile displacement_list_${frame} fast_${frame} greater 93.5

    #create new multibodies comprised of clusters of particles with high displacements that remain neighbors after a displacement time index of ${frame}
    persistent_neighbors ${frame} vneighbors p_neighbors_${frame} per_neighbors centroid
    list fast_${frame}

    #report out size statistics on these persistent neighbor clusters, including 4 moments of the distribution
    size_statistics data/neigh/p_neighbors_${frame}.dat p_neighbors_${frame} 4

    #create new multibodies comprised of clusters of particles with high displacements that remain neighbors after a displacement time index of ${frame}
    persistent_neighbors ${frame} dneighbors p_dneighbors_${frame} per_neighbors centroid
    list fast_${frame}

    #report out size statistics on these persistent neighbor clusters, including 4 moments of the distribution
    size_statistics data/neigh/p_dneighbors_${frame}.dat p_dneighbors_${frame} 4

    delete_trajectory_list p_dneighbors_${frame}
    delete_trajectory_list p_neighbors_${frame}
    delete_trajectory_list fast_${frame}
    delete_valuelist displacement_list_${frame}

end

Despite including these lines before the loop increments $frame to the next value, I still have to increase the amount of memory for the job to complete.
Am I missing something here? Is there another trajectory_list/value_list that I am not deleting?
Is any of this documented anywhere?

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Fields

No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions