# Note: # Try to avoid module level import statements here to reduce # import time during CLI execution import sys from ase.cli.main import CLIError template_help = """ Without argument, looks for ~/.ase/template.py. Otherwise, expects the comma separated list of the fields to include in their left-to-right order. Optionally, specify the lexicographical sort hierarchy (0 is outermost sort) and if the sort should be ascending or descending (1 or -1). By default, sorting is descending, which makes sense for most things except index (and rank, but one can just sort by the thing which is ranked to get ascending ranks). * example: ase diff start.cif stop.cif --template * i:0:1,el,dx,dy,dz,d,rd possible fields: * i: index * dx,dy,dz,d: displacement/displacement components * dfx,dfy,dfz,df: difference force/force components * afx,afy,afz,af: average force/force components * p1x,p1y,p1z,p: first image positions/position components * p2x,p2y,p2z,p: second image positions/position components * f1x,f1y,f1z,f: first image forces/force components * f2x,f2y,f2z,f: second image forces/force components * an: atomic number * el: atomic element * t: atom tag * r: the rank of that atom with respect to the column It is possible to change formatters in the template file.""" class CLICommand: """Print differences between atoms/calculations. Supports taking differences between different calculation runs of the same system as well as neighboring geometric images for one calculation run of a system. As part of a difference table or as a standalone display table, fields for non-difference quantities of image 1 and image 2 are also provided. See the --template-help for the formatting exposed in the CLI. More customization requires changing the input arguments to the Table initialization and/or editing the templates file. """ @staticmethod def add_arguments(parser): add = parser.add_argument add('file', help="""Possible file entries are * 2 non-trajectory files: difference between them * 1 trajectory file: difference between consecutive images * 2 trajectory files: difference between corresponding image numbers * 1 trajectory file followed by hyphen-minus (ASCII 45): for display Note deltas are defined as 2 - 1. Use [FILE]@[SLICE] to select images. """, nargs='+') add('-r', '--rank-order', metavar='FIELD', nargs='?', const='d', type=str, help="""Order atoms by rank, see --template-help for possible fields. The default value, when specified, is d. When not specified, ordering is the same as that provided by the generator. For hierarchical sorting, see template.""") add('-c', '--calculator-outputs', action="store_true", help="display calculator outputs of forces and energy") add('--max-lines', metavar='N', type=int, help="show only so many lines (atoms) in each table " ", useful if rank ordering") add('-t', '--template', metavar='TEMPLATE', nargs='?', const='rc', help="""See --template-help for the help on this option.""") add('--template-help', help="""Prints the help for the template file. Usage `ase diff - --template-help`""", action="store_true") add('-s', '--summary-functions', metavar='SUMFUNCS', nargs='?', help="""Specify the summary functions. Possible values are `rmsd` and `dE`. Comma separate more than one summary function.""") add('--log-file', metavar='LOGFILE', help="print table to file") add('--as-csv', action="store_true", help="output table in csv format") add('--precision', metavar='PREC', default=2, type=int, help="precision used in both display and sorting") @staticmethod def run(args, parser): import io if args.template_help: print(template_help) return encoding = 'utf-8' if args.log_file is None: out = io.TextIOWrapper(sys.stdout.buffer, encoding=encoding) else: out = open(args.log_file, 'w', encoding=encoding) with out: CLICommand.diff(args, out) @staticmethod def diff(args, out): from ase.cli.template import ( Table, TableFormat, energy_delta, field_specs_on_conditions, rmsd, slice_split, summary_functions_on_conditions, ) from ase.io import read if args.template is None: field_specs = field_specs_on_conditions( args.calculator_outputs, args.rank_order) else: field_specs = args.template.split(',') if not args.calculator_outputs: for field_spec in field_specs: if 'f' in field_spec: raise CLIError( "field requiring calculation outputs " "without --calculator-outputs") if args.summary_functions is None: summary_functions = summary_functions_on_conditions( args.calculator_outputs) else: summary_functions_dct = { 'rmsd': rmsd, 'dE': energy_delta} summary_functions = args.summary_functions.split(',') if not args.calculator_outputs: for sf in summary_functions: if sf == 'dE': raise CLIError( "summary function requiring calculation outputs " "without --calculator-outputs") summary_functions = [summary_functions_dct[i] for i in summary_functions] have_two_files = len(args.file) == 2 file1 = args.file[0] actual_filename, index = slice_split(file1) atoms1 = read(actual_filename, index) natoms1 = len(atoms1) if have_two_files: if args.file[1] == '-': atoms2 = atoms1 def header_fmt(c): return f'image # {c}' else: file2 = args.file[1] actual_filename, index = slice_split(file2) atoms2 = read(actual_filename, index) natoms2 = len(atoms2) same_length = natoms1 == natoms2 one_l_one = natoms1 == 1 or natoms2 == 1 if not same_length and not one_l_one: raise CLIError( "Trajectory files are not the same length " "and both > 1\n{}!={}".format( natoms1, natoms2)) elif not same_length and one_l_one: print( "One file contains one image " "and the other multiple images,\n" "assuming you want to compare all images " "with one reference image") if natoms1 > natoms2: atoms2 = natoms1 * atoms2 else: atoms1 = natoms2 * atoms1 def header_fmt(c): return f'sys-ref image # {c}' else: def header_fmt(c): return f'sys2-sys1 image # {c}' else: atoms2 = atoms1.copy() atoms1 = atoms1[:-1] atoms2 = atoms2[1:] natoms2 = natoms1 = natoms1 - 1 def header_fmt(c): return f'images {c + 1}-{c}' natoms = natoms1 # = natoms2 output = '' tableformat = TableFormat(precision=args.precision, columnwidth=7 + args.precision) table = Table( field_specs, max_lines=args.max_lines, tableformat=tableformat, summary_functions=summary_functions) for counter in range(natoms): table.title = header_fmt(counter) output += table.make(atoms1[counter], atoms2[counter], csv=args.as_csv) + '\n' print(output, file=out)