import subprocess
import time
import matplotlib.pyplot as plt
import numpy as np

frams_cli_path = r'C:\Users\kapis\Desktop\PUT\magisterka\semestr2\Komosiński\LAB\Framstics\Framsticks50rc13\frams.exe'
commands = [f'"{frams_cli_path}" "expdef standardMY" "Simulator.init();" "im case_study_experiment_state8.expt" "im case_study_paremeters3.sim" "Math.randomize();" "Simulator.start();" "while (ExpState.totaltestedcr<600) Simulator.step();" "sa koniec_%%N.expt" "-q"',
f'"{frams_cli_path}" "expdef standardMY" "Simulator.init();" "im case_study_experiment_state8.expt" "im case_study_paremeters3_slow.sim" "Math.randomize();" "Simulator.start();" "while (ExpState.totaltestedcr<600) Simulator.step();" "sa koniec_%%N.expt" "-q"']

def invoke_calculations(command):
    p = subprocess.Popen(command, stdout=subprocess.PIPE, shell=True)
    output, err = p.communicate()
    p.wait()
    return output


def prepare_significant_lines(output):
    lines = output.decode().splitlines()
    significant_lines = [line for line in lines if "[LOG]" in line]
    return significant_lines


def extract_fitnesses(lines):
    return [line.split(' ')[-2] for line in lines]


def process():
    times = []
    best_fitnesses = []
    iteration_fitnesses = []

    for index, command in enumerate(commands):
        times.append([])
        best_fitnesses.append([])
        iteration_fitnesses.append([])
        for _ in range(10):
            start = time.clock()
            output = invoke_calculations(command)
            end = time.clock()

            elapsed = end - start
            times[index].append(np.array(elapsed).astype(np.float))

            significant_lines = prepare_significant_lines(output)
            fitnesses = extract_fitnesses(significant_lines)
            iteration_fitnesses[index].append(np.array(fitnesses).astype(np.float))
            max_fitness = max(fitnesses)
            best_fitnesses[index].append(np.array(max_fitness).astype(np.float))

    fig, ax = plt.subplots()
    bp1 = ax.boxplot(best_fitnesses, positions=[1, 4], notch=True, widths=0.35,
                     patch_artist=True, boxprops=dict(facecolor="C0"))
    bp2 = ax.boxplot(times, positions=[2, 5], notch=True, widths=0.35,
                     patch_artist=True, boxprops=dict(facecolor="C2"))

    ax.legend([bp1["boxes"][0], bp2["boxes"][0]], ['Fitness', 'Time'], loc='upper right')

    ax.set_xlim(0, 6)
    plt.savefig('boxplots_fitnesses_times.svg')

    fig, ax = plt.subplots()
    bp1 = ax.boxplot(best_fitnesses[0], positions=[1], notch=True, widths=0.70,
                     patch_artist=True, boxprops=dict(facecolor="C0"))
    bp2 = ax.boxplot(best_fitnesses[1], positions=[2], notch=True, widths=0.70,
                     patch_artist=True, boxprops=dict(facecolor="C2"))

    ax.legend([bp1["boxes"][0], bp2["boxes"][0]], ['parm1', 'parm2_slow'], loc='upper right')

    ax.set_xlim(0, 3)
    plt.savefig('boxplots_fitnesses.svg')

    fig, ax = plt.subplots()
    bp1 = ax.boxplot(times[0], positions=[1], notch=True, widths=0.70,
                     patch_artist=True, boxprops=dict(facecolor="C0"))
    bp2 = ax.boxplot(times[1], positions=[2], notch=True, widths=0.70,
                     patch_artist=True, boxprops=dict(facecolor="C2"))

    ax.legend([bp1["boxes"][0], bp2["boxes"][0]], ['parm1', 'parm2_slow'], loc='upper right')

    ax.set_xlim(0, 3)
    plt.savefig('boxplots_times.svg')
    plt.show()

    for index, parametrization in enumerate(iteration_fitnesses):
        for iteration in parametrization:
            plt.plot(iteration, '')
            plt.ylabel('Fitness')
            plt.xlabel('Iteration')
        plt.savefig(f'lines{index}.svg')
        plt.show()

if __name__ == '__main__':
    process()