backgammon/plot.py

import os
import pandas as pd
from datetime import datetime
import csv
import datetime
import matplotlib.pyplot as plt
import matplotlib.ticker as mtick
import matplotlib.dates as mdates

train_headers = ['timestamp', 'eps_train', 'eps_trained_session', 'sum', 'mean']
eval_headers  = ['timestamp', 'method', 'eps_train', 'eval_eps_used', 'sum', 'mean']
bench_headers = ['method', 'sample_count', 'i', 'time', 'sum', 'mean']

model_path = 'models'

def plot_bench(data_path):
    df = pd.read_csv(data_path, sep=";",
                     names=bench_headers, index_col=[0,1,2])
    for method_label in df.index.levels[0]:
        df_prime = df[['mean']].loc[method_label].unstack().T
        plot = df_prime.plot.box()
        plot.set_title("Evaluation variance, {}".format(method_label))
        plot.set_xlabel("Sample count")
        plot.set_ylabel("Mean score")
        plt.show(plot.figure)

        # for later use:
        variances = df_prime.var()
        print(variances)

        del df_prime, plot, variances

def dataframes(model_name):
    def df_timestamp_to_datetime(df):
        df['timestamp'] = df['timestamp'].map(lambda t: datetime.datetime.fromtimestamp(t))
        return df
    
    log_path = os.path.join(model_path, model_name, 'logs')
    raw_dfs = [ pd.read_csv(os.path.join(log_path, 'eval.log'), sep=';', names=eval_headers),
                pd.read_csv(os.path.join(log_path, 'train.log'), sep=';', names=train_headers) ]
    dfs = [ df_timestamp_to_datetime(df) for df in raw_dfs ]
    dataframes = {
        'eval': dfs[0],
        'train': dfs[1] 
    }
    return dataframes


if __name__ == '__main__':
    fig, (ax1, ax2) = plt.subplots(2, 1, sharex=True, sharey=True)

    #plt.ion()
    ax1.set_title('Mean over episodes')
    ax2.set_xlabel('Episodes trained')
    ax1.set_ylabel('Points-per-game')
    ax1.grid(True)
    ax2.grid(True)

    #ax.set_xlim(left=0)
    ax1.set_ylim([-2, 2])

    df = dataframes('tesauro-5')['eval']

    print(df)

    dumbeval_df = df.query("method == 'dumbeval'")
    pubeval_df  = df.query("method == 'pubeval'")

    def plot_eval(axis, label, df, c):
        x = df['eps_train']
        y = df['mean']
    
        axis.scatter(x, y, label=label, c=c, marker="x")

    plot_eval(ax1, "dumbeval", dumbeval_df, [[1, 0.5, 0]])
    plot_eval(ax2, "pubeval",  pubeval_df,  [[0, 0.5, 1]])

    ax1.legend()
    ax2.legend()

    
    plt.show()