MayFes.py

MayFes.py

# -*- coding: utf-8 -*-
import pandas as pd
import datetime
import sys
from collections import defaultdict
import matplotlib.pyplot as plt
import argparse

# define constants
MAX_MEMBER = 6 # excluding band master
AVAILABLE = 0 # available is 0, unavailable is 1

# For analysis
def compute_num_of_band():
	hist_y = []
	hist_x = []
	for day in range(0, final_day-initial_day+1):
		for hour in range(first_h, last_h):
			count = 0
			for i in range(len(data)):
				count += data.ix[i]["{}-{}".format(day+1, hour+1)]

			print "{}-{}".format(day+1, hour+1), "AVAILABLE: {}".format(total_band_num-count)
			hist_x.append("{}-{}".format(day+1, hour+1))
			hist_y.append(total_band_num-count)

	plt.barh([i for i in range(len(hist_y))], hist_y)
	plt.yticks([i for i in range(len(hist_y))], hist_x)
	plt.title("The number of band that can perform")
	plt.show()

# For analysis
def compute_band_hist():
	plt.hist(data.total)
	plt.show()
	# sys.exit()

def compute_mean_age(data):
	mean_age = []
	import re
	r = re.compile(r"\d+")
	for i in range(len(data)):
		n_member = 0
		age = 0.0
		for j in range(MAX_MEMBER):
			age_str = data.ix[i]["member{}".format(j+1)]
			if pd.isnull(age_str):
				continue

			res = r.search(age_str)
			if res == None:
				continue

			age += float(int(res.group(0)))
			n_member += 1

		age /= float(n_member)
		mean_age.append(age)

	return mean_age

def main():
	parser = argparse.ArgumentParser()
	parser.add_argument("input_csv",
		 help="Input csv file.")
	# parse options
	parser.add_argument("--initial_day",
		type=int,
		help="initial day.")
	parser.add_argument("--final_day",
		type=int,
		help="final day.")
	parser.add_argument("--initial_hour",
		type=int,
		help="initial hour.")
	parser.add_argument("--last_hour",
		type=int,
		help="last hour.")
	parser.add_argument("--rest_time",
		default=15,
		type=int,
		help="last hour.")
	args = parser.parse_args()

	data = pd.read_csv(args.input_csv)
	initial_day = args.initial_day
	final_day = args.final_day
	rest_time = args.rest_time
	initial_hour = args.initial_hour
	last_hour = args.last_hour

	# results = []
	total_band_num = data.shape[0]

	# add mean of ki
	mean_age = compute_mean_age(data)
	data["mean_age"] = pd.Series(mean_age, index=data.index)

	# print data
	now_time = datetime.datetime(2015, 5, initial_day, initial_hour, 0, 0)
	while True:

		# Error check
		if now_time > datetime.datetime(now_time.year, now_time.month, now_time.day, last_hour-1, 59, 0):
			now_time = datetime.datetime(now_time.year, now_time.month, now_time.day+1, 9, 0, 0)

		if now_time.day > final_day:
			break

		# Check whether to insert break.
		if now_time.hour == 12 and now_time.minute > 15:
			print "Break!", now_time
			now_time += datetime.timedelta(seconds=45*60)
			continue

		if now_time.hour % 2 == 0 and now_time.minute > 50:
			print "Break!", now_time
			now_time += datetime.timedelta(seconds=rest_time*60)
			continue

		# results.append(defaultdict(str))

		# select candidates of band
		initial_candidate = data[data["{}-{}".format(now_time.day - initial_day + 1, now_time.hour + 1)]==AVAILABLE]
		if len(initial_candidate) == 0:
			print "Break!", now_time
			now_time += datetime.timedelta(seconds=rest_time*60)
			continue

		# Filter by available time slot
		max_total = initial_candidate.total.max()
		candidate_by_total = initial_candidate[initial_candidate.total==max_total]

		# Filter by mean age
		max_mean_age = candidate_by_total.mean_age.max()
		candidate_by_mean_age = candidate_by_total[candidate_by_total.mean_age==max_mean_age]

		band = candidate_by_total.iloc[0]
		band_index = candidate_by_total.index[0]
		data = data.drop(band_index)
		print now_time, band.band_name

		now_time += datetime.timedelta(seconds=band.required_time*60)

	print
	print "Band not listed:"
	for band in data.band_name.values:
		print band

if __name__ == '__main__':
	main()