gurobi_log_plot.py

import re
import argparse

import pandas as pd
import matplotlib.pyplot as plt


integer = "[0-9]+"
double = "[0-9]+\.*[0-9]*"
_double = "[0-9]+\.[0-9]+([eE][+-]?[0-9]+)"

line_log_pattern = re.compile(
    f"\s*(?P<expl>{integer})\s+(?P<unexpl>{integer})\s+(?P<obj>{double})\s+(?P<depth>{integer})\s+(?P<intinf>{integer})\s+(?P<incumbent>{double})\s+(?P<bestbd>{double})\s+(?P<gap>{double})%\s+(?P<itnode>{double})\s+(?P<time>{integer})s"
)

result_log_pattern1 = re.compile(
    f"Explored (?P<expl>{integer}) nodes \((?P<simplex_iter>{integer}) simplex iterations\) in (?P<time>{double}) seconds \((?P<work_units>{double}) work units\)"
)

result_log_pattern2 = re.compile(
    f"Best objective (?P<obj>{_double}), best bound (?P<bestbd>{_double}), gap (?P<gap>{double})%"
)


def main(log_file):

    logs = list()

    for line in open(log_file, "r"):
        line = line.strip()

        if m := line_log_pattern.match(line):
            logs.append(m.groupdict())
        elif m := result_log_pattern1.match(line):
            result1 = m.groupdict()
        elif m := result_log_pattern2.match(line):
            result2 = m.groupdict()
            objective_value = float(result2["obj"])

    df = pd.DataFrame(logs).astype(float)  # cast from str to float
    objective_value = float(objective_value)  # cast from str to float

    fig, ax = plt.subplots()

    # plot incumbent and lower bound
    df.plot(x="time", y="incumbent", ax=ax)
    df.plot(x="time", y="bestbd", ax=ax)

    # plot objective value
    x_min, x_max = 0, float(df.tail(1).time)
    ax.hlines(
        objective_value, x_min, x_max, color="red", linestyles="--", label="optimal"
    )

    # some settings
    ax.set_xlabel("Time [s]")
    ax.set_ylabel("Objective value")
    ax.grid("--")
    ax.legend()


if __name__ == "__main__":

    parser = argparse.ArgumentParser()
    parser.add_argument(
        "gurobi_log_file",
        type=str,
    )
    args = parser.parse_args()
    main(args.gurobi_log_file)