Coverage for intelligence_toolkit/tests/unit/anonymize_case_data/test_synthesizability_statistics.py: 100%

1# Copyright (c) 2024 Microsoft Corporation. All rights reserved. 

2# Licensed under the MIT license. See LICENSE file in the project. 

3 

4import pytest 

5from intelligence_toolkit.anonymize_case_data.synthesizability_statistics import ( 

6 SynthesizabilityStatistics, 

7) 

8 

9 

10def test_synthesizability_statistics_initialization(): 

11 stats = SynthesizabilityStatistics( 

12 num_cols=5, 

13 overall_att_count=100, 

14 possible_combinations=1000, 

15 possible_combinations_per_row=10.5, 

16 mean_vals_per_record=3.2, 

17 max_combinations_per_record=9.2, 

18 excess_combinations_ratio=1.14, 

19 ) 

20 

21 assert stats.num_cols == 5 

22 assert stats.overall_att_count == 100 

23 assert stats.possible_combinations == 1000 

24 assert stats.possible_combinations_per_row == 10.5 

25 assert stats.mean_vals_per_record == 3.2 

26 assert stats.max_combinations_per_record == 9.2 

27 assert stats.excess_combinations_ratio == 1.14 

28 

29 

30def test_synthesizability_statistics_repr(): 

31 stats = SynthesizabilityStatistics( 

32 num_cols=3, 

33 overall_att_count=50, 

34 possible_combinations=500, 

35 possible_combinations_per_row=5.0, 

36 mean_vals_per_record=2.5, 

37 max_combinations_per_record=5.7, 

38 excess_combinations_ratio=0.88, 

39 ) 

40 

41 repr_str = repr(stats) 

42 

43 assert "SynthesizabilityStatistics" in repr_str 

44 assert "num_cols=3" in repr_str 

45 assert "overall_att_count=50" in repr_str 

46 assert "possible_combinations=500" in repr_str 

47 assert "possible_combinations_per_row=5.0" in repr_str 

48 assert "mean_vals_per_record=2.5" in repr_str 

49 assert "max_combinations_per_record=5.7" in repr_str 

50 assert "excess_combinations_ratio=0.88" in repr_str 

51 

52 

53def test_synthesizability_statistics_zero_values(): 

54 stats = SynthesizabilityStatistics( 

55 num_cols=0, 

56 overall_att_count=0, 

57 possible_combinations=0, 

58 possible_combinations_per_row=0.0, 

59 mean_vals_per_record=0.0, 

60 max_combinations_per_record=0.0, 

61 excess_combinations_ratio=0.0, 

62 ) 

63 

64 assert stats.num_cols == 0 

65 assert stats.overall_att_count == 0 

66 assert stats.possible_combinations == 0 

67 assert stats.possible_combinations_per_row == 0.0 

68 assert stats.mean_vals_per_record == 0.0 

69 assert stats.max_combinations_per_record == 0.0 

70 assert stats.excess_combinations_ratio == 0.0