Coverage for C:\src\imod-python\imod\msw\coupler_mapping.py: 100%

59 statements  

« prev     ^ index     » next       coverage.py v7.5.1, created at 2024-05-08 14:15 +0200

1from typing import Optional 

2 

3import numpy as np 

4import pandas as pd 

5import xarray as xr 

6 

7from imod.mf6.dis import StructuredDiscretization 

8from imod.mf6.wel import WellDisStructured 

9from imod.msw.fixed_format import VariableMetaData 

10from imod.msw.pkgbase import MetaSwapPackage 

11 

12 

13class CouplerMapping(MetaSwapPackage): 

14 """ 

15 This contains the data to connect MODFLOW 6 cells to MetaSWAP svats. 

16 

17 This class is responsible for the file `mod2svat.inp`. It also includes 

18 connection to wells. 

19 

20 Parameters 

21 ---------- 

22 modflow_dis: StructuredDiscretization 

23 Modflow 6 structured discretization 

24 well: WellDisStructured (optional) 

25 If given, this parameter describes sprinkling of SVAT units from MODFLOW 

26 cells. 

27 """ 

28 

29 _file_name = "mod2svat.inp" 

30 _metadata_dict = { 

31 "mod_id": VariableMetaData(10, 1, 9999999, int), 

32 "free": VariableMetaData(2, None, None, str), 

33 "svat": VariableMetaData(10, 1, 9999999, int), 

34 "layer": VariableMetaData(5, 0, 9999, int), 

35 } 

36 

37 _with_subunit = ("mod_id",) 

38 _without_subunit = () 

39 _to_fill = ("free",) 

40 

41 def __init__( 

42 self, 

43 modflow_dis: StructuredDiscretization, 

44 well: Optional[WellDisStructured] = None, 

45 ): 

46 super().__init__() 

47 

48 self.well = well 

49 # Test if equal or larger than 1, to ignore idomain == -1 as well. Don't 

50 # assign to self.dataset, as grid extent might differ from svat when 

51 # MetaSWAP only covers part of the Modflow grid domain. 

52 self.idomain_active = modflow_dis["idomain"] >= 1 

53 

54 def _create_mod_id_rch(self, svat): 

55 """ 

56 Create modflow indices for the recharge layer, which is where 

57 infiltration will take place. 

58 """ 

59 self.dataset["mod_id"] = xr.full_like(svat, fill_value=0, dtype=np.int64) 

60 n_subunit = svat["subunit"].size 

61 idomain_top_active = self.idomain_active.sel(layer=1, drop=True) 

62 

63 n_mod_top = idomain_top_active.sum() 

64 

65 # idomain does not have a subunit dimension, so tile for n_subunits 

66 mod_id_1d = np.tile(np.arange(1, n_mod_top + 1), (n_subunit, 1)) 

67 

68 self.dataset["mod_id"].values[:, idomain_top_active.values] = mod_id_1d 

69 

70 def _render(self, file, index, svat): 

71 self._create_mod_id_rch(svat) 

72 # package check only possible after calling _create_mod_id_rch 

73 self._pkgcheck() 

74 

75 data_dict = {"svat": svat.values.ravel()[index]} 

76 

77 data_dict["layer"] = np.full_like(data_dict["svat"], 1) 

78 

79 for var in self._with_subunit: 

80 data_dict[var] = self._index_da(self.dataset[var], index) 

81 

82 # Get well values 

83 if self.well: 

84 mod_id_well, svat_well, layer_well = self._create_well_id(svat) 

85 data_dict["mod_id"] = np.append(mod_id_well, data_dict["mod_id"]) 

86 data_dict["svat"] = np.append(svat_well, data_dict["svat"]) 

87 data_dict["layer"] = np.append(layer_well, data_dict["layer"]) 

88 

89 for var in self._to_fill: 

90 data_dict[var] = "" 

91 

92 dataframe = pd.DataFrame( 

93 data=data_dict, columns=list(self._metadata_dict.keys()) 

94 ) 

95 

96 self._check_range(dataframe) 

97 

98 return self.write_dataframe_fixed_width(file, dataframe) 

99 

100 def _create_well_id(self, svat): 

101 """ 

102 Get modflow indices, svats, and layer number for the wells 

103 """ 

104 n_subunit = svat["subunit"].size 

105 

106 # Convert to Python's 0-based index 

107 well_row = self.well["row"] - 1 

108 well_column = self.well["column"] - 1 

109 well_layer = self.well["layer"] - 1 

110 

111 n_mod = self.idomain_active.sum() 

112 mod_id = xr.full_like(self.idomain_active, 0, dtype=np.int64) 

113 mod_id.values[self.idomain_active.values] = np.arange(1, n_mod + 1) 

114 

115 well_mod_id = mod_id[well_layer, well_row, well_column] 

116 well_mod_id = np.tile(well_mod_id, (n_subunit, 1)) 

117 

118 well_svat = svat.values[:, well_row, well_column] 

119 

120 well_active = well_svat != 0 

121 

122 well_svat_1d = well_svat[well_active] 

123 well_mod_id_1d = well_mod_id[well_active] 

124 

125 # Tile well_layers for each subunit 

126 layer = np.tile(well_layer + 1, (n_subunit, 1)) 

127 layer_1d = layer[well_active] 

128 

129 return (well_mod_id_1d, well_svat_1d, layer_1d)