def run_pipeline():
output_fname = util.get_daily_lineup_out_fname()
now = get_curr_date()
if not os.path.exists(output_fname):
team_depth_charts = depth_chart.scrape_depth_chart()
games = schedule_parser.get_games_on_date(datetime.datetime.strftime(now, "%m/%d/%Y"))
daily_lineup = []
for game in games:
home = game["home"]
away = game["away"]
date = game["date"]
home_depth_data = team_depth_charts[home]
away_depth_data = team_depth_charts[away]
daily_lineup_data = {
"date": date,
"home": home,
"away": away,
"home_depth": home_depth_data,
"away_depth": away_depth_data,
}
daily_lineup.append(daily_lineup_data)
# now_str = datetime.datetime.strftime(now, '%m-%d-%Y')
# output_filename = "pipeline_out/daily_lineups_" + now_str + ".json"
f = open(output_fname, "w+")
f.write(json.dumps(daily_lineup, sort_keys=True, indent=4))
f.close()
def get_last_n_stats(self, player_name, n=5, query_stats=None): if not query_stats: query_stats = self.DEFAULT_QUERY_STATS today = get_curr_date() select_stats = '' # @TODO: Design a way so that you can easily tell what type a column is in your database # Right now, we're explicitly checking for the date so that we don't convert it to float # representation for stat in query_stats: select_stats += stat + ', ' #Truncate extra comma and space at end, if there was at least one entry for query stats if len(query_stats) > 0: select_stats = select_stats[0:-2] template = 'SELECT ' + select_stats + ' FROM player_stats WHERE name = ? and date <= ? ORDER BY date DESC LIMIT ? ' results = self.cursor.execute(template, [player_name, today, n]).fetchall() total = 0 for row in results: total += row[-1] print total/5 return results
def get_last_n_averaged_stats(self, player_name, n=5, query_stats=None): if not query_stats: query_stats = self.DEFAULT_QUERY_STATS today = get_curr_date() last_n_stats = np.array(self.get_last_n_stats(player_name, n, query_stats)) indices = [i for i in range(len(query_stats)) if query_stats[i] != 'Date' and query_stats[i] != 'opp_team'] last_n_stats = last_n_stats[:, indices].astype(float) avgs = list(np.mean(last_n_stats, axis=0)) avgs = [round(avg, 2) for avg in avgs] print(avgs) return avgs """
def refresh(self): self.today = get_curr_date() self.now = get_curr_time() self.load_data() self.analyze_games()
"""
Constructs a mapping where the keys are player names, and values are their Fanduel-unique attributes,
such as salary data and average fantasy point values.
"""
from util import get_curr_date
import util
from datetime import datetime
import pandas
import json
import os
now = datetime.strftime(get_curr_date(), '%m-%d-%Y')
def run_pipeline():
pipeline_in = util.get_fd_pipe_in_fname()
pipeline_out = util.get_fd_pipe_out_fname()
if os.path.exists(pipeline_out):
f = open(pipeline_out, 'r')
data = json.loads(f.read())
f.close()
return data
player_mapping = {}
dataframe = pandas.read_csv(pipeline_in)
nrows = len(dataframe)
for c in range(nrows):
