def analytics(**kwargs):
# python cf.py --analytics --readfile output12.txt --stat t
# python cf.py --analytics --readfile output12.txt --stat win_player --pct_analytics 1
# python cf.py --runs 300 --strat_player (1,) --strat_me --strat_you
# python cf.py --runs 300 --strat_fork (1,)
# python cf.py --analytics --readfile output91.txt --stat win_player --pct_analytics 1
log = Log(noisy=False)
data = log.load_from_file(filename=args['readfile'])
stat_str = kwargs.get('stat', args['stat'])
stat = log.get_stat(data, stat=stat_str)
#filter the Nones
stat_v = [val for val in stat if val is not None]
#analytics
avg = float(sum(stat_v)) / float(len(stat_v))
_min, _max = min(stat_v), max(stat_v)
print 'Var: ', str(stat_str.upper())
print 'N: ', str(len(stat_v))
print 'Avg: ', str(avg)[:5]
print 'Min: ', str(_min), ' Max: ', str(_max)
if len(args['pct_analytics']) > 0:
match = str(args['pct_analytics'])
num_match = [str(x) for x in stat_v].count(match)
pct_match = 100.0 * float(num_match) / float(len(stat_v))
print 'PCT of ', stat_str, ' = ', match, ' : ', str(pct_match)[:5]
def batch():
log = Log(noisy=False, noisy_win=args["noisy_gamewin"])
board = Board(BOARD_WIDTH, BOARD_HEIGHT)
log.batch_board_params(board)
#AI-Rules
strat = KnownRules(players=strat_players,
c3me=c3me,
c3you=c3you,
forkme=forkme)
log.batch_strat_params(strat)
for game_i in range(int(args['runs'])):
log.game_start(game_i=game_i)
play = Play(board=board, state=copy.deepcopy(INIT_STATE))
for i in range(BOARD_WIDTH * BOARD_HEIGHT + 1):
#ACTION-SPACE
ap = play.available_plays()
if len(ap) < 1:
log.game_tie()
break
#DECISION-ACTION
ret = strat.strategize(play, log, board)
log.strat_played(ret, play, noisy=args["noisy_strat"])
playcol = ret[0] if int(ret[0]) > -1 else random.sample(ap, 1)[0]
log.game_play(playcol, play)
play.make_play(playcol)
#EVAL-PAYOFF-FUNCTION
if play.check_win(log=log): #this writes to log too
log.game_win_print(win_state=play.state,
noisy=args["noisy_gamewin"])
break
log.game_end()
log.end_runs()
from os import path
from matplotlib import pyplot as plt
from probscale import plot_pos, probplot
import numpy as np
import math
from src.utils import chalk, loadSeries, loadCSV, runFunctions, Log, Filer, Series, Format
from src.utils import constants
from src.Q1 import solveQuestion1, Gamma
from src.utils.estimators import Estimator
from src.utils.frequency import Frequency
name = "Questão 2"
log = Log().instance(prepend=chalk.cyan("Q2"))
curr_location = path.join('.', 'src', 'Q2')
A1 = loadSeries("aluno3_A1.csv")
freq_ranges = [3, 6, 9, 12, 15, 18, 21, 24]
def probPlots(verbose=False):
from scipy import stats
for method in constants.EstimatorMethod.options:
estimators = solveQuestion1(method)
for estimator in estimators:
# Create a new figure
plt.figure()
xlog = log.instance().indentMore()
def __init__(self):
super(History, self).__init__()
self.logs = Log()