Output

There are a number of files created during the execution of k-ART. Details on the content of each file is given at the end of this manual. Here is a simple description.

provides a global summary of all simulation details, including parameters, code version, the date of the compilation and a number of used in the simulation as well as the basic simulation time as a function of event.
Sortieproc.xx files are the log files for each core running the simulation (labelled from 0 to N-1, where N is the number of cores being used). These give explicit details about each step of the simulation. The MASTER is always labelled 0 (sortieproc.0) and all other cores are WORKERS.
Output files where data for further analysis have a .dat extension:
1. Energies.dat lets you see the number of KMC step, CPU times, simulated times, activated barrier executed. It is contains the basic information about the evolution;
2. Diffusion.dat provides information on the mean squares displacement with the initial configuration as a reference;
3. selec_ev.dat provides information on the event chosen by the code at each KMC step.
4. Gen_ev.dat and Spec_ev.dat give respectively information on creation of generics and specifics events.
Files Topologies and topos.list are a exhaustive list of all the the topology keys observed during the simulation.
allconfs give us the possibility to visualize the spatial evolution atoms of the system during the simulation.(allconf_defect for example let us visualize only atoms in a noncristaline topology)
Multiple event_list_conf_x.dat files provide a list of all the topologies and events present in configuration x.
Finally, events.uft is an unformatted fortran file listing all the event of the simulation and the folder EVENT_DIR which describe a readable equivalent of events.uft

Detailed content of output files

Files event_list_conf_$n$ located in the EVLIST_DIR provides a exhaustive overview of the available transition of the simulation that leads to the KMC step $n$ for the local minimum $n-1$. It counts 11 columns

(1) TypeId : Atomic type (be careful type does not mean chemical element i.e. Si1 Si2 are to different type of Si).

(2) AtomId: Id of the Atom.

(3) TopoId: Topology of the local initial minimum.

(4) EventId : Generic Event Id.

(5) Spec_id: Specific Event Id.

(6) barrier: Refine or reconverge or clone energy barrier (in eV) see last logical column.

(7) inv_bar: Refine or reconverge or clone energy of inverse barrier (in eV) see last logical column.

(8) asy_ener: Difference between barrier and inv_barrier

(9) inisad_dr: displacement between saddle and initial configuration.

(10) inifin_dr: displacement between final and initial configuration.

(11) noname logical: if true means the event was refined or reconverged otherwise it was cloned.
The file Energies.dat provides a general overview of the evolution of the simulation. It counts 8 columns with each row providing the information for each minimum of energy $E_{Min}^n$ reached at step $n$):

(1) Number of KMC step $n$.

(2) CPU time.

(3) Old minimum Energy $E_{Min}^{n-1}$.

(4) New minimum energy $E_{Min}^n$.

(5) Activation Energy or Energy barrier (or saddle energy minus the energy of the initial minimum) that is selected by kART from the list of all available events from this configuration, i.e., $E^n_a=E^n_{Sad}-E_{Min}^{n-1}$.

(6) Total rate sum of all the transition rates $\sum_i r_i$ that are extracted from the activation energy base on the transition state theory $r_i = \tau_0 e^{-E ^i_a/k_BT}$, at each step $n$. If a value of 0.0 is found, it means that the mean rate method (MRM) has been used for leaving a basin (in this case, the minimum energy shown in column 3 will not correspond to the value of column 4 at the previous step because internal dynamics in a basin is lost).

(7) Time step (e.g., $\Delta t=-\ln \mu/ \sum_i r_i$, where $\mu$ is a random number in the interval $(0,1)$).

(8) Simulated time (it is the accumulative sum of the previous time steps).

(9) Selec Ev, id of the generic event selected for transition.

(10) Init Topo, id of the initial topology.

(11) Sad Topo, id of the saddle topology.

(12) Fin Topo, id of the final topology.
The file Diffusion.dat gives the information on the mean squares displacement with respect to the the initial configuration. It is displayed info on 4 columns:

(1) Elapsed Time, time step $t_n$.

(2) Sqr Displ, total square displacement, $\hbox{{\it SD}}(t_n)=\sum_{i=1}^N \left(r_i(t_n)-r_i(0)\right)^2$, of $N$ atoms at $t_n$. If several atoms, the column is splitted into several subcolumns where the first is the Total SD of the system, the second is SD of the atoms of type 1, third is the SD of atoms of type 2, etc (the type order is the same as in the input file with the coord).

(3) KMC step, is the step $n$.

(4) Event delr, sum of the displacement from initial, $r^{ini}$, to final position, $r^{final}$, of the atoms that moved during the transition, that is, delr $=\sqrt{\sum_{i=1}^N (r_i^{final}-r_i^{ini})^2}$. The sum usually corresponds to the displacement of the atom that moves most during the step; as the others remain at their initial positions (for example, if an interstitial, this value approximately corresponds to the displacement of the interstitial).
The file selec_ev.dat gives us information on the event chosen by the code at each KMC step.\

(1) CPU time: real time used to get that step.

(2) selcBarrier: activation energy or energy barrier, that is $E^n_a=E^n_{Sad}-E_{Min}^{n-1}$.

(3) KMC: number of KMC step $n$.

(4) EVENT: the generic event selected during the step (same number found in file Energies.dat ).

(5) SPEC_EVENT: the generic event executed during the step. For example, if the generic EVENT selected is 887554, there is a subset of specific events from which one is executed, for instance, the number 9 thus we identify it as 887554_9. In contrast to generic events, specific events take into account local deformations and stresses of the particular configuration at that KMC step (see section [9] sec:generating_events_art_part {reference-type=”ref” reference=”sec:generating_events_art_part”}).

(6) SELEC_ATOM: main atom over the event is executed. It is the atom that advances most during the step.

(7) timestep: as before $\Delta t=-\ln \mu/ \sum_i r_i$, where $\mu$ is a random number in the interval $(0,1)$.

(8) SimulT: (Simulated time), it is the accumulative sum of the previous time steps.

(9) Basin ID: The basin ID indicates to thing if we are in a BMRM bassin (value $\ne0$) or not (value $0$). the non-value where we are in a BMRM bassin increment according to the number of different bassin visited. if this value as the same value for several following step this means that we are still in the same BMRM bassin.

(10) Basin thresh: maximum height of barrier and inverse barrier for an event to be considered inside a basin used in the step as set by MIN_SIG_BARRIER.
Files Gen_ev.dat give us information on creation of generics events. To avoid ambiguity, if one read the line at the $i^{th}$ KMCstep this line represents the statistic about the transition from $i^{th}-1$ to $i^{th}$. It has the following information:

(1) KMC: number of KMC step $n$.

(2) Number of Generic event search: this is the total amount of ARTn search per KMC steps.

(3) Number of saddle attempts: this is the total number of saddle attempts (for each ARTn search there is 10 independent attempts to find a saddle point) this column includes the failed and succeed attempts. Again per KMC steps.

(4) Number of saddle found: the is the total number of saddle found per KMC steps.

(5) Generic event found Excluding inverse event: count all successful generation of Generic events

(6) Generic event found Including inverse event: count all successful generation of Generic events except the inverse Generic event.

(7) New Generic event found Excluding inverse event: only count the one that are not in catalogue.

(8) New Generic event found Including inverse event: only count the one that are not in catalogue.

(9) Tot GenEv Incl InvEv: Total number of distinct Generic event in catalogue.

(10) Force eval: this is the number of force evaluations for ARTn part of the code per KMC steps.
The file Spec_ev.dat give us information on creation of specifics events.To avoid ambiguity, if one read the line at the $i^{th}$ KMCstep this line represents the statistic about the transition from $(i-1)^{th}$ to $i^{th}$. It has the following information:

(1) KMC: number of KMC step $n$.

(2) # GenEv to Consider: This is the number of Generic event to refine into Specific be careful this does not count the multiplicity of atoms sharing this Generic event.

(3) # of SpecEv to Refine: This is the number of Specific event to refine a saddle point contrary to the previous one this does count the multiplicity of atoms since it is a Spec event.

(4) # of SpecEv to Reconverge: This is the number of Specific event to reconverge a saddle point we do reconvergence of SpecEv when a previous SpecEv exists in the actual configuration.

(5) # of SpecEv to be Cloned: this is the number of event that should be cloned.

(6) Spec Ev found: this is the number of Spec event found.

(7) New Spec Ev: this is the number of new spec event found.

(8) Tot SpecEv: this is the total number of distinct Specific event in catalogue.

(9) Thres cloning: this is the energy barrier threshold at which we start cloning Generic events.

(10) Ev Cloned: this is the number of cloned events

(11) Tot Ev: this is the total number of available events that leads to the KMC step $n$.
Files Topologies and topos.list are a exhaustive list of all the the topology keys observed during the simulation. Accordingly to the NAUTY the file Topologies basically shows:

(1) topo_labels: topology label or hash key.

(2) id1, id2, id3: the integers we use as input to a hash function to generate a unique topology number (hash key).

(3) topo cutoff: it is the length-cutoff used by default to link two atoms (see input parameters for topology in KMC.sh file).
The file topo_stat.dat has information about topologies at each step. It gives the following data:

(1) CPU time.

(2) currTOPO: is the current total number of topologies in the system. Equivalent to “known_tc” variable in code and in sortiproc.0 file, it only increases if the variable STATISTICS is set to True.

(3) activeTOPO: number of topologies that are active in the system (equivalent to variable “active_tc” in sortiproc.0 file and in the code, it only increases if the variable STATISTICS is set to True).

(4) newTOPO: the number of new topologies found at each step. The name of the variable in the code is “new_tc” and only increases if the variable STATISTICS is set to True.

(5) searchedTOPO: number of topologies total (all steps). The name of variable in the code is “searched_tc” and only increases if the variable STATISTICS is set to True.

(6) CumulTOPO: total number of topologies found in the simulation. The name of variable in the code is “cumul_tc” and only increases if the variable STATISTICS is set to True.

(7) diffTOPO: number of difficult topologies found during the step.

(8) KMC: step.

(9) SimulT: total time.
allconf files give us the possibility to visualize the spatial evolution atoms of the system during the simulation. This file is given in XYZ format that can be visualized using for example, VMD, OVITO, VESTA, Rasmol, etc, programs. (allconf_defect for example let us visualize only atoms in a no-cristaline topology)
Multiple event_list_conf_n.dat files are found in the dir EVLIST_DIR. Each file gives a list of the topologies and events found at each step that is used to create the catalog for choosing the next step (for example, event_list_conf_8.dat has the list used to choose the step 8, an other way to say it is that it has a list that leads from step 7 to 8). Each file gives the following data:

(1) AtomId: id of the atom.

(2) TopoId: id of corresponding topology of the AtomId.

(3) EventId: id of one of the events associated to the the topology TopoId.

(4) Spec_id: id for the specific event, which is used mostly for debugging or recovering detailed history of the kinetics. If Spec_id $= 1$ it is a generic event otherwise it is a specific event.

(5) Barrier: barrier energy, $E_a=E_{sad}-E_{ini}$, corresponding to EventId.

(6) Inv_bar: inverse barrier $E_i=E_{sad}-E_{final}$.

(7) Asy_ener: final minus initial minimum energies: $E_{final}-E_{ini}$.

(8) Inisad_dr: displacement from initial to saddle position of the atoms that moved most during the event finding, that is $\sqrt{\sum_{i=1}^N (r_i^{sad}-r_i^{ini})^2}$.

(9) Inifin_dr: same value as “Event delr” in Diffusion.dat or displacement from initial to final position of the atoms that moved most during the event finding, that is $\sqrt{\sum_{i=1}^N (r_i^{final}-r_i^{ini})^2}$.
File events.uft, is a unformatted fortran file listing all the events of the simulation contained in the folder EVENT_DIR (this folder contains the same information as events.uft, but human readable and given in different files). events.uft is useful for importing the list of events when we want to continue a previous simulation.
File topos.list has technical details of topologies of the system. Data is printed in two lines, the first has 6 fields: field one is the unique topology label or “hash key” created from a set of three topology indexes given by NAUTY (fields from two to four); the field five is the MAX_TOPO_CUTOFF and the field six is a flag (T or F), for identifying a topology as a difficult or not. The second line has five fields: the first is this_topo%count,this_topo%attempts,this_topo%failed_attempts, this_topo%ncluster, this_topo%primary_topo_id [ ….STILL REMAINS TO EXPLAIN ….]{style=”color: red”}
File output.dat prints a list of parameters and their setting values for the particular simulation. This file should be self explicative.
This file KMC_log.txt is a global resume of all the detail (physics parameters but also the version number of the code used even the date of the compilation) used in a simulation.
Finally, sortieproc.n files are the log files for each core running the simulation (labelled from 0 to N-1, where N is the number of cores being used). These give explicit details about each step of the simulation. The MASTER is always labelled 0 (sortieproc.0) and all other cores are SLAVES.