Coverage for NeuralTSNE/NeuralTSNE/Plotter/plot.py: 97%

1from typing import List 

2import matplotlib.pyplot as plt 

3import numpy as np 

4 

5 

6def plot( 

7 data: np.ndarray, 

8 labels: np.ndarray | None, 

9 step: int, 

10 marker_size: int, 

11 alpha: float, 

12 are_neural_labels: bool = False, 

13 img_file: str | None = None, 

14 kwargs: dict | None = None, 

15) -> None: 

16 """ 

17 Plot t-SNE results. 

18 

19 Parameters 

20 ---------- 

21 `data` : `np.ndarray` 

22 t-SNE data to be plotted. 

23 `labels` : `np.ndarray`, optional 

24 Labels corresponding to the data points. 

25 `step` : `int` 

26 Step size for subsampling the data. 

27 `marker_size` : `int` 

28 Marker size for the scatter plot. 

29 `alpha` : `float` 

30 Alpha value for transparency in the scatter plot. 

31 `are_neural_labels` : `bool`, optional 

32 Flag indicating whether the labels are neural network predictions. 

33 `img_file` : `str`, optional 

34 File path to save the plot as an image. 

35 `**kwargs` : `dict`, optional 

36 Additional keyword arguments. 

37 

38 Important 

39 --------- 

40 The following additional keyword arguments are available: 

41 

42 `file_step` : `int`, optional 

43 Step size for subsampling labels. Defaults to `1`. 

44 

45 Note 

46 ---- 

47 This function plots the t-SNE results with scatter plot, allowing customization of various plot parameters. 

48 """ 

49 if kwargs is None: 

50 kwargs = {} 

51 f_step = kwargs.get("file_step", 1) 

52 

53 plt.subplots(1, 1) 

54 

55 ( 

56 plt.scatter( 

57 data[::step, 0], 

58 data[::step, 1], 

59 marker_size, 

60 alpha=alpha, 

61 marker=".", 

62 ) 

63 if labels is None 

64 else plt.scatter( 

65 data[::step, 0], 

66 data[::step, 1], 

67 marker_size, 

68 labels[:: f_step * step] if not are_neural_labels else labels[::step], 

69 alpha=alpha, 

70 marker=".", 

71 ) 

72 ) 

73 

74 plt.ylabel("t-SNE 2") 

75 plt.xlabel("t-SNE 1") 

76 

77 if img_file: 77 ↛ 80line 77 didn't jump to line 80 because the condition on line 77 was always true

78 new_name = img_file 

79 plt.savefig(new_name) 

80 plt.show() 

81 

82 

83def plot_from_file( 

84 file: str, 

85 labels_file: str, 

86 columns: List[int], 

87 step: int, 

88 marker_size: int, 

89 alpha: float, 

90 are_neural_labels: bool = False, 

91) -> None: 

92 """ 

93 Plot t-SNE results from file. 

94 

95 Parameters 

96 ---------- 

97 `file` : `str` 

98 File path containing t-SNE data. 

99 `labels_file` : `str` 

100 File path containing labels data. 

101 `columns` : `List[int]` 

102 Column indices to load from the labels file. 

103 `step` : `int` 

104 Step size for subsampling the data. 

105 `marker_size` : `int` 

106 Marker size for the scatter plot. 

107 `alpha` : `float` 

108 Alpha value for transparency in the scatter plot. 

109 `are_neural_labels` : `bool`, optional 

110 Flag indicating whether the labels are neural network predictions. 

111 

112 Note 

113 ---- 

114 This function reads t-SNE data and labels from files, applies subsampling, and plots the results using the `plot` function. 

115 """ 

116 data = None 

117 file_step = None 

118 

119 with open(file, "r") as f: 

120 file_step = int(f.readline()) 

121 data = np.loadtxt(f) 

122 

123 labels = None 

124 if labels_file: 

125 with open(labels_file, "r") as f: 

126 labels = np.loadtxt(f, usecols=columns, dtype="int") 

127 data = data[: len(labels)] 

128 

129 plot( 

130 data, 

131 labels, 

132 step, 

133 marker_size, 

134 alpha, 

135 are_neural_labels, 

136 file.rsplit(".", 1)[0] + ".png", 

137 {"file_step": file_step}, 

138 )