File size: 6,594 Bytes
8e5930e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
import copy
import eyekit as ek
import numpy as np
import pandas as pd
from PIL import Image


MEASURES_DICT = {
    "number_of_fixations": [],
    "initial_fixation_duration": [],
    "first_of_many_duration": [],
    "total_fixation_duration": [],
    "gaze_duration": [],
    "go_past_duration": [],
    "second_pass_duration": [],
    "initial_landing_position": [],
    "initial_landing_distance": [],
    "landing_distances": [],
    "number_of_regressions_in": [],
}


def get_fix_seq_and_text_block(
    dffix,
    trial,
    x_txt_start=None,
    y_txt_start=None,
    font_face="Courier New",
    font_size=None,
    line_height=None,
    use_corrected_fixations=True,
    correction_algo="warp",
):
    if use_corrected_fixations and correction_algo is not None:
        fixations_tuples = [
            (
                (x[1]["x"], x[1][f"y_{correction_algo}"], x[1]["corrected_start_time"], x[1]["corrected_end_time"])
                if x[1]["corrected_start_time"] < x[1]["corrected_end_time"]
                else (x[1]["x"], x[1]["y"], x[1]["corrected_start_time"], x[1]["corrected_end_time"] + 1)
            )
            for x in dffix.iterrows()
        ]
    else:
        fixations_tuples = [
            (
                (x[1]["x"], x[1]["y"], x[1]["corrected_start_time"], x[1]["corrected_end_time"])
                if x[1]["corrected_start_time"] < x[1]["corrected_end_time"]
                else (x[1]["x"], x[1]["y"], x[1]["corrected_start_time"], x[1]["corrected_end_time"] + 1)
            )
            for x in dffix.iterrows()
        ]

    try:
        fixation_sequence = ek.FixationSequence(fixations_tuples)
    except Exception as e:
        print(e)
        print(f"Creating fixation failed for {trial['trial_id']} {trial['filename']}")
        return dffix

    if "display_coords" in trial:
        display_coords = trial["display_coords"]
    else:
        display_coords = (0, 0, 1920, 1080)
    screen_size = ((display_coords[2] - display_coords[0]), (display_coords[3] - display_coords[1]))

    y_diffs = np.unique(trial["line_heights"])
    if len(y_diffs) == 1:
        y_diff = y_diffs[0]
    else:
        y_diff = np.min(y_diffs)
    chars_list = trial["chars_list"]
    max_line = int(chars_list[-1]["assigned_line"])
    words_on_lines = {x: [] for x in range(int(max_line) + 1)}
    [words_on_lines[x["assigned_line"]].append(x["char"]) for x in chars_list]
    sentence_list = ["".join([s for s in v]) for idx, v in words_on_lines.items()]

    if x_txt_start is None:
        x_txt_start = float(chars_list[0]["char_xmin"])
    if y_txt_start is None:
        y_txt_start = float(chars_list[0]["char_ymax"])

    if font_face is None and "font" in trial:
        font_face = trial["font"]
    elif font_face is None:
        font_face = "DejaVu Sans Mono"

    if font_size is None and "font_size" in trial:
        font_size = trial["font_size"]
    elif font_size is None:
        font_size = float(y_diff * 0.333)  # pixel to point conversion
    if line_height is None:
        line_height = float(y_diff)
    textblock = ek.TextBlock(
        sentence_list,
        position=(float(x_txt_start), float(y_txt_start)),
        font_face=font_face,
        line_height=line_height,
        font_size=font_size,
        anchor="left",
        align="left",
    )

    # eyekit_plot(textblock, fixation_sequence, screen_size)
    ek.io.save(fixation_sequence, f'results/fixation_sequence_eyekit_{trial["trial_id"]}.json', compress=False)
    ek.io.save(textblock, f'results/textblock_eyekit_{trial["trial_id"]}.json', compress=False)

    return fixation_sequence, textblock, screen_size


def eyekit_plot(textblock, fixation_sequence, screen_size):
    img = ek.vis.Image(*screen_size)
    img.draw_text_block(textblock)
    for word in textblock.words():
        img.draw_rectangle(word, color="hotpink")
    img.draw_fixation_sequence(fixation_sequence)
    img.save("temp_eyekit_img.png", crop_margin=200)
    img_png = Image.open("temp_eyekit_img.png")
    return img_png


def plot_with_measure(textblock, fixation_sequence, screen_size, measure, use_characters=False):

    eyekitplot_img = eyekit_plot(textblock, fixation_sequence, screen_size)
    eyekitplot_img = ek.vis.Image(*screen_size)
    eyekitplot_img.draw_text_block(textblock)
    if use_characters:
        measure_results = getattr(ek.measure, measure)(textblock.characters(), fixation_sequence)
        enum = textblock.characters()
    else:
        measure_results = getattr(ek.measure, measure)(textblock.words(), fixation_sequence)
        enum = textblock.words()
    for word in enum:
        eyekitplot_img.draw_rectangle(word, color="lightseagreen")
        x = word.onset
        y = word.y_br - 3
        label = f"{measure_results[word.id]}"
        eyekitplot_img.draw_annotation((x, y), label, color="lightseagreen", font_face="Arial bold", font_size=15)
    eyekitplot_img.draw_fixation_sequence(fixation_sequence, color="gray")
    eyekitplot_img.save("multiline_passage_piccol.png", crop_margin=100)
    img_png = Image.open("multiline_passage_piccol.png")
    return img_png


def get_eyekit_measures(_txt, _seq, get_char_measures=False):
    measures = copy.deepcopy(MEASURES_DICT)
    words = []
    for w in _txt.words():
        words.append(w.text)
        for m in measures.keys():
            measures[m].append(getattr(ek.measure, m)(w, _seq))
    word_measures_df = pd.DataFrame(measures)
    word_measures_df["word_number"] = np.arange(0, len(words))
    word_measures_df["word"] = words

    first_column = word_measures_df.pop("word")
    word_measures_df.insert(0, "word", first_column)
    first_column = word_measures_df.pop("word_number")
    word_measures_df.insert(0, "word_number", first_column)

    if get_char_measures:
        measures = copy.deepcopy(MEASURES_DICT)

        characters = []
        for c in _txt.characters():
            characters.append(c.text)
            for m in measures.keys():
                measures[m].append(getattr(ek.measure, m)(c, _seq))
        character_measures_df = pd.DataFrame(measures)
        character_measures_df["char_number"] = np.arange(0, len(characters))
        character_measures_df["character"] = characters

        first_column = character_measures_df.pop("character")
        character_measures_df.insert(0, "character", first_column)
        first_column = character_measures_df.pop("char_number")
        character_measures_df.insert(0, "char_number", first_column)
    else:
        character_measures_df = None
    return word_measures_df, character_measures_df