plot.py

from pathlib import Path

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from matplotlib import style


def main():
    #     scale = [0, 0.01, 0.02, 0.03, 0.04, 0.05]
    #     nsfiair = [96.05, 97.11, 97.46, 97.46, 97.58, 97.56]

    #     vae = [98.5, 98.5, 98.5, 96.9, 96.3, 93.1]

    #     style.use("seaborn")
    #     fig, plot = plt.subplots(figsize=(6, 4), dpi=200)
    #     plot.plot(scale, nsfiair, label="NoSINN", marker="v")
    #     plot.plot(scale, vae, label="CVAE", marker="s")
    #     plot.set_xlabel(r"$\sigma$")
    #     plot.set_ylabel("Accuracy")
    #     plot.set_title("CMNIST")
    #     plot.set_ylim(90, 100)
    #     plot.legend()
    #     fig.savefig(Path(".") / "cmnist.pdf")
    #     plt.show()

    # df = pd.read_csv("/its/home/mb715/DOcuments/Fairness/FINN/results/adult/adult_cae.csv")
    # relevant_columns = [
    #     "Mix_fact", "Accuracy",
    #     "prob_pos_sex_Male_0-sex_Male_1",
    #     "TPR_sex_Male_0-sex_Male_1"
    # ]
    # df = df[relevant_columns]
    # df = df.rename(columns={
    #     "Mix_fact": "Mixing factor",
    #     "prob_pos_sex_Male_0-sex_Male_1": "DP",
    #     "TPR_sex_Male_0-sex_Male_1": "EO"
    # })
    # latex = df.to_latex(
    #     index=False,
    #     float_format="{:0.4f}".format)
    # print(latex)

    df = pd.read_csv(
        "/its/home/mb715/Documents/Fairness/FINN/results/celeba/celeba_naive_baseline_pred_y_25epochs.csv"
    )
    relevant_columns = ["Accuracy", "prob_pos_sens_0-sens_1", "TPR_sens_0-sens_1"]
    df = df[relevant_columns]
    df = df.rename(
        columns={
            # "Mix_fact": "Mixing factor",
            "prob_pos_sens_0-sens_1": "DP",
            "TPR_sens_0-sens_1": "EO",
        }
    )
    print(len(df))
    df.insert(0, "Mixing factor", np.arange(0, 0.9, 0.1), True)
    latex = df.to_latex(index=False, float_format="{:0.4f}".format)
    print(latex)


if __name__ == "__main__":
    main()