Coverage for /builds/ase/ase/ase/symbols.py: 96.08%

1# fmt: off 

2 

3import collections.abc 

4import numbers 

5import warnings 

6from typing import Dict, Iterator, List, Sequence, Set, Union 

7 

8import numpy as np 

9 

10from ase.data import atomic_numbers, chemical_symbols 

11from ase.formula import Formula 

12 

13Integers = Union[Sequence[int], np.ndarray] 

14 

15 

16def string2symbols(s: str) -> List[str]: 

17 """Convert string to list of chemical symbols.""" 

18 return list(Formula(s)) 

19 

20 

21def symbols2numbers(symbols) -> List[int]: 

22 if isinstance(symbols, str): 

23 symbols = string2symbols(symbols) 

24 numbers = [] 

25 for s in symbols: 

26 if isinstance(s, str): 

27 numbers.append(atomic_numbers[s]) 

28 else: 

29 numbers.append(int(s)) 

30 return numbers 

31 

32 

33class Symbols(collections.abc.Sequence): 

34 """A sequence of chemical symbols. 

35 

36 ``atoms.symbols`` is a :class:`ase.symbols.Symbols` object. This 

37 object works like an editable view of ``atoms.numbers``, except 

38 its elements are manipulated as strings. 

39 

40 Examples: 

41 

42 >>> from ase.build import molecule 

43 >>> atoms = molecule('CH3CH2OH') 

44 >>> atoms.symbols 

45 Symbols('C2OH6') 

46 >>> atoms.symbols[:3] 

47 Symbols('C2O') 

48 >>> atoms.symbols == 'H' # doctest: +ELLIPSIS 

49 array([False, False, False, True, True, True, True, True, True]...) 

50 >>> atoms.symbols[-3:] = 'Pu' 

51 >>> atoms.symbols 

52 Symbols('C2OH3Pu3') 

53 >>> atoms.symbols[3:6] = 'Mo2U' 

54 >>> atoms.symbols 

55 Symbols('C2OMo2UPu3') 

56 >>> atoms.symbols.formula 

57 Formula('C2OMo2UPu3') 

58 

59 The :class:`ase.formula.Formula` object is useful for extended 

60 formatting options and analysis. 

61 

62 """ 

63 

64 def __init__(self, numbers) -> None: 

65 self.numbers = np.asarray(numbers, int) 

66 

67 @classmethod 

68 def fromsymbols(cls, symbols) -> 'Symbols': 

69 numbers = symbols2numbers(symbols) 

70 return cls(np.array(numbers)) 

71 

72 @property 

73 def formula(self) -> Formula: 

74 """Formula object.""" 

75 string = Formula.from_list(self).format('reduce') 

76 return Formula(string) 

77 

78 def __getitem__(self, key) -> Union['Symbols', str]: 

79 num = self.numbers[key] 

80 if isinstance(num, numbers.Integral): 

81 return chemical_symbols[num] 

82 return Symbols(num) 

83 

84 def __iter__(self) -> Iterator[str]: 

85 for num in self.numbers: 

86 yield chemical_symbols[num] 

87 

88 def __setitem__(self, key, value) -> None: 

89 numbers = symbols2numbers(value) 

90 if len(numbers) == 1: 

91 self.numbers[key] = numbers[0] 

92 else: 

93 self.numbers[key] = numbers 

94 

95 def __len__(self) -> int: 

96 return len(self.numbers) 

97 

98 def __str__(self) -> str: 

99 return self.get_chemical_formula('reduce') 

100 

101 def __repr__(self) -> str: 

102 return f'Symbols(\'{self}\')' 

103 

104 def __eq__(self, obj) -> bool: 

105 if not hasattr(obj, '__len__'): 

106 return False 

107 

108 try: 

109 symbols = Symbols.fromsymbols(obj) 

110 except Exception: 

111 # Typically this would happen if obj cannot be converged to 

112 # atomic numbers. 

113 return False 

114 return self.numbers == symbols.numbers 

115 

116 def get_chemical_formula( 

117 self, 

118 mode: str = 'hill', 

119 empirical: bool = False, 

120 ) -> str: 

121 """Get chemical formula. 

122 

123 See documentation of ase.atoms.Atoms.get_chemical_formula().""" 

124 # XXX Delegate the work to the Formula object! 

125 if mode in ('reduce', 'all') and empirical: 

126 warnings.warn("Empirical chemical formula not available " 

127 "for mode '{}'".format(mode)) 

128 

129 if len(self) == 0: 

130 return '' 

131 

132 numbers = self.numbers 

133 

134 if mode == 'reduce': 

135 n = len(numbers) 

136 changes = np.concatenate(([0], np.arange(1, n)[numbers[1:] != 

137 numbers[:-1]])) 

138 symbols = [chemical_symbols[e] for e in numbers[changes]] 

139 counts = np.append(changes[1:], n) - changes 

140 

141 tokens = [] 

142 for s, c in zip(symbols, counts): 

143 tokens.append(s) 

144 if c > 1: 

145 tokens.append(str(c)) 

146 formula = ''.join(tokens) 

147 elif mode == 'all': 

148 formula = ''.join([chemical_symbols[n] for n in numbers]) 

149 else: 

150 symbols = [chemical_symbols[Z] for Z in numbers] 

151 f = Formula('', _tree=[(symbols, 1)]) 

152 if empirical: 

153 f, _ = f.reduce() 

154 if mode in {'hill', 'metal'}: 

155 formula = f.format(mode) 

156 else: 

157 raise ValueError( 

158 "Use mode = 'all', 'reduce', 'hill' or 'metal'.") 

159 

160 return formula 

161 

162 def search(self, symbols) -> Integers: 

163 """Return the indices of elements with given symbol or symbols.""" 

164 numbers = set(symbols2numbers(symbols)) 

165 indices = [i for i, number in enumerate(self.numbers) 

166 if number in numbers] 

167 return np.array(indices, int) 

168 

169 def species(self) -> Set[str]: 

170 """Return unique symbols as a set.""" 

171 return set(self) 

172 

173 def indices(self) -> Dict[str, Integers]: 

174 """Return dictionary mapping each unique symbol to indices. 

175 

176 >>> from ase.build import molecule 

177 >>> atoms = molecule('CH3CH2OH') 

178 >>> atoms.symbols.indices() 

179 {'C': array([0, 1]), 'O': array([2]), 'H': array([3, 4, 5, 6, 7, 8])} 

180 

181 """ 

182 dct: Dict[str, List[int]] = {} 

183 for i, symbol in enumerate(self): 

184 dct.setdefault(symbol, []).append(i) 

185 return {key: np.array(value, int) for key, value in dct.items()} 

186 

187 def species_indices(self) -> Sequence[int]: 

188 """Return the indices of each atom within their individual species. 

189 

190 >>> from ase import Atoms 

191 >>> atoms = Atoms('CH3CH2OH') 

192 >>> atoms.symbols.species_indices() 

193 [0, 0, 1, 2, 1, 3, 4, 0, 5] 

194 

195 ^ ^ ^ ^ ^ ^ ^ ^ ^ 

196 C H H H C H H O H 

197 

198 """ 

199 

200 counts: Dict[str, int] = {} 

201 result = [] 

202 for i, n in enumerate(self.numbers): 

203 counts[n] = counts.get(n, -1) + 1 

204 result.append(counts[n]) 

205 

206 return result