ReactionMechanismGenerator/ARC: ARC 1.1.0

ARC Version 1.1.0

ARC - Automated Rate Calculator is a software for automating electronic structure calculations relevant for chemical kinetic modeling. ARC has many advanced options, yet at its core it is simple: it accepts 2D graph representations of chemical species (e.g., SMILES or adjacency lists), and automatically executes, tracks, and processes relevant electronic structure jobs on user-defined servers. The principal outputs of ARC are thermodynamic properties (H, S, Cp) and high-pressure limit kinetic rate coefficients for the defined species and reactions. (Note that automating transition states is still in progress).

• New features:

  • Project name validity check #58
  • Added Free Rotors #59
  • Report the git head in the log file #63
  • Determine rotor symmetry using a cyclic energy scan list #66
  • Improve SGE submit file #68
  • Improved restart #64
  • Parse xyz for a species from file #70
  • Troubleshoot conformer jobs if they all fail #77
  • Changed rotor inconsistency_ab #78
  • Don't crush ARC if it couldn't troubleshoot a job #79
  • Make xyz to 2D more robust (catch errors) #80
  • Check isomorphism for conformers #84
  • Improve submit files #81
  • Troubleshoot rotor scans #85
  • Generate resonance structures before checking molecule isomorphism #87
  • Standardize xyz #88
  • Copy the frequency output file for a TS #89
  • Save the generated conformers in the species geometry folder #90
  • Allow nonisomorphic 2d #91
  • Improved management and report of job time #94
  • Don't use applyAtomEnergyCorrections for kinetics #95
  • Fix to xyz/2D loading and atom reordering in ARCSpecies #96
  • Preserve multiplicity throughout Species representations #99
  • Archive log and restart files #97
  • Run and save molecular orbitals #102
  • Allow ARCSpecies to be initialized with a conformer list file #104
  • Set default max stepsize for optimization in Gaussian #108
  • Save unconverged species (labels) #98
  • Use the Gaussian check file #100
  • Allow birad singlet to be defined and run as unrestricted #114
  • Update determine_qm_software to new Arkane #111
  • Calculating Lennard-Jones coefficients #117
  • Allow the user to define adaptive levels of theory (by heavy atoms) #119
  • Set the 1d_rotors settings directive to True by default #121
  • Updated submit scripts #123
  • Allow defining a TS species simply with .xyz guesses #124
  • Save conformer jobs in the restart file #125
  • Parse Molpro frequencies #127
  • Run ARC on a server #128
  • Activate the job_types flags for opt, freq, and sp #132
  • Modifications to how the model chemistry is treated to be consistent with Arkane #134
  • Organized job memory handling #137
  • Consider the requested job types for species already defined with xyz #139
  • Added a ZPE/freq scaling factor script to utils #136
  • Added a function to save an input file (and not run ARC on spot) #140
  • Added some Job methods and reorganized init #142
  • Added NIST to kinetics plots (off by default) #146
  • A new way of thinking about conformers #143
  • Improved ESS troubleshooting #147
  • Adapt Processor to the new Arkane output #148

• Documentation:

  • Added ARC's documentation #152

• Bug fixes:

  • Plot kinetics even if no RMG reaction matches #62
  • Treat same scan point at 0 and 360 deg correctly #67
  • Fix bug with loading molecule statmech during Rate jobs #73
  • If statement in model chemistry #101
  • Don't lose track of jobs in 'qw' status on SGE #103
  • Don't overwrite self.mol if self.mol_from_xyz is None #106
  • Remove extra parentheses in gaussian input file #120
  • Don't run a fine opt in Molpro #122
  • Set the memory per cpu slightly higher than the ESS memory #144
  • File compatibility with MAC OS (skip .DS_Store files) #145
  • Improved ARC's sleeping habits when experiencing connection errors #149
  • Minor conformers/xyz files fixes #150
  • Correcly move the Arkane YAML file #151

• Miscellaneous:

  • #109:
    • Added examples to .gitignore
    • Allow users to specify the number of cpu's to use per server
    • Download additional info re failed jobs (out.txt, err.txt, slurm.out)
    • Don't run orbitals jobs by default
    • Make servers in ess_settings a list
    • Allow users to specify a levels_ess dictionary associating levels of theory (or partial phrases of methods or basis sets) with an ESS
    • Added ess_settings to settings.py instead of the API
    • Tests: functions in main
    • Additional minor fixes
  • #112:
    • Load ess_settings correctly from dict
    • Improved species logging in project info file
    • Don't plot kinetics if it is None
    • Improved error message if rxn energetics are problematic
    • Added min_list() to Scheduler
    • Improve converter xyz functions
    • Improved xyz handeling in Species (xyz can now be a list)
  • #133:
    • Updated isomorphic species check method name in rmgdb
    • Corrected local server check
    • Updated ESS check function name in iPython notebook
  • #141:
    • Added an iPy notebook with an xyz to SMILES script
    • Don't load the RMG database if not necessary

• New dependencies:

  • scikit-learn #116