cell_viability_paper/figure_feature_contribution.py at master · KISRDevelopment/cell_viability_paper · GitHub

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
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import scipy.stats
import utils.stats
import seaborn as sns
import json
import os

plot_cfg = {
    "tick_label_size" : 60,
    "xlabel_size" : 60,
    "ylabel_size" : 65,
    "border_size" : 10,
    "bar_border_size" : 2.5,
    "bar_label_size" : 65,
    "stars_label_size" : 48,
    "annot_size" : 72,
    "max_cm_classes" : 4,
    "max_bars" : 7,
    "legend_label_size" : 42
}

ALPHA = 0.05

plt.rcParams["font.family"] = "Liberation Serif"
plt.rcParams["font.weight"] = "bold"

def generate_figures(spec, path, output_path):

    df = create_plot_df(spec, path)
    plot_df(spec, df, output_path)

def create_plot_df(spec, path):

    results_path = os.path.join(path, spec['models'][0]['name'], "results.json")

    cv_results = load_results(results_path)

    # reference baccs
    baccs = [r['bacc'] for r in cv_results]

    rows = []
    for model in spec['models'][1:]:
        results_path = os.path.join(path, model['name'], "results.json")

        cv_results = load_results(results_path)

        for i, r in enumerate(cv_results):
            rows.append({
                "model" : model['title'],
                "bacc" : r['bacc'],
                "drop_perc" : (baccs[i] - r['bacc']) * 100 / baccs[i],
                "color" : model['color'],
                "star_color" : model.get('star_color', model['color'])
            })

    df = pd.DataFrame(rows)
    print(df)
    return df

def load_results(path):
    with open(path, 'r') as f:
        results = json.load(f)['results']

    return sorted(results, key=lambda r: r['split_id'])

def plot_df(spec, df, output_path):
    n_models = len(spec['models']) - 1
    models = spec['models'][1:]

    num_comparisons = n_models * (n_models - 1) / 2
    adjusted_alpha = ALPHA / num_comparisons

    ylim = spec['ylim']

    g = sns.catplot(x="model",
        y="drop_perc",
        data=df,
        kind="bar", ci="sd",
        height=10,
        aspect=spec.get('aspect', 1),
        palette=[m['color'] for m in models],
        edgecolor='black',
        errwidth=5,
        errcolor='black',
        linewidth=plot_cfg["bar_border_size"],
        saturation=1)

    ax = g.ax

    # plot pvalues
    offset_val = 0.01 * ylim[1]
    incr_val = 0.04 * ylim[1]

    for i, model in enumerate(models):
        a_ix = df['model'] == model['title']
        a_bacc = df[a_ix]['bacc']
        a_drop = df[a_ix]['drop_perc']

        yoffset = offset_val

        for j in range(i+1, n_models):
            b_ix = df['model'] == models[j]['title']
            b_bacc = df[b_ix]['bacc']

            statistic, pvalue = scipy.stats.ttest_rel(a_bacc, b_bacc)

            if pvalue < adjusted_alpha:
                stars = '*' * utils.stats.compute_stars(pvalue, adjusted_alpha)

                target_color = models[j].get('star_color', models[j]['color'])

                ax.text(i,np.mean(a_drop) + np.std(a_drop) + yoffset, stars, color=target_color, ha="center", va="center", weight='bold', fontsize=plot_cfg['stars_label_size'])
                yoffset += incr_val

    ax.yaxis.set_tick_params(labelsize=plot_cfg["tick_label_size"])
    ax.xaxis.set_tick_params(labelsize=plot_cfg["tick_label_size"], rotation=90)
    ax.set_xlabel('')
    ax.set_ylabel('% Drop', fontsize=plot_cfg["ylabel_size"], weight='bold')


    ax.yaxis.set_tick_params(length=10, width=1, which='both')
    ax.set_ylim(ylim)
    plt.setp(ax.spines.values(), linewidth=plot_cfg["border_size"], color='black')

    plt.savefig(output_path, bbox_inches='tight')

if __name__ == "__main__":

    spec = {
        "models": [
            {
                "title": "S-MN",
                "color": "#FF0000",
                "name" : "smf"
            },
            {
                "title": "No LID",
                "color": "magenta",
                "name" : "smf_no_lid"
            },
            {
                "title": "No sGO",
                "color": "cyan",
                "name" : "smf_no_sgo"
            },
            {
                "title": "No Homology",
                "color": "#d62728",
                "name" : "smf_no_redundancy"
            },
        ],
        "ylim" : [0,30],
        "aspect" : 1
    }

    generate_figures(spec, "../results/exp_mn_feature_contribution",
        "../results/exp_mn_feature_contribution/bacc_drop.png")