Skip to content

Commit

Permalink
arbitrage: add email serializer
Browse files Browse the repository at this point in the history
  • Loading branch information
@zhehaowang
zhehaowang committed Jun 15, 2019
1 parent 5fd29e8 commit 6686f49
Show file tree
Hide file tree
Showing 2 changed files with 212 additions and 22 deletions.
224 changes: 205 additions & 19 deletions src/arbitrage_locator/arbitrage.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,10 +5,19 @@
import sys
import json
import csv
import pytz

import pprint
import datetime

import argparse

# Import smtplib for the actual sending function
import smtplib

# Import the email modules we'll need
from email.mime.text import MIMEText

pp = pprint.PrettyPrinter()

def sanitize_key(key):
Expand Down Expand Up @@ -42,7 +51,7 @@ def read_files(fc_file, stockx_file):
style_id = row[2]
name = row[0]
url = 'https://stockx.com/' + row[1]
shoe_size = row[3]
shoe_size = sanitize_size(row[3])
best_bid = float(row[4])
best_ask = float(row[5])
sx_volume_last_72 = int(row[6])
Expand Down Expand Up @@ -97,7 +106,7 @@ def match_items(fc_prices, sx_prices):
sx_item = sx_prices[style_id][sx_size]
if not style_id in matches:
matches[style_id] = {
size: {
sx_size: {
"fc": fc_item,
"sx": sx_item
}}
Expand All @@ -120,6 +129,16 @@ def find_margin(matches):
"sx_xxx": ...,
"fc_yyy": ...}}
"""
# us to china
# du_shipping_fee = 20
# du_commission_rate = 0.095

# china to us
# du_batch_size = 3

# du_batch_fee = 106
# 90 - 5kg + 16 / kg

sx_threshold = 50
sx_bid_commission = 13.95
fc_commission_rate = 0.2
Expand Down Expand Up @@ -149,10 +168,11 @@ def find_margin(matches):

fc_url_with_size = item['fc']['url'] + '?size=' + size
match_item = {
# 'style_id': style_id,
'shoe_size': size,
'style_id': style_id,
'name': item['sx']['name'],
'shoe_size': sanitize_size(size),

'fc_px': item['fc']['px'],
'fc_px': float(item['fc']['px'].replace(',', '')),
'fc_url': fc_url_with_size,

'sx_best_bid': item['sx']['best_bid'],
Expand Down Expand Up @@ -195,25 +215,191 @@ def annotate_transaction_history(sx_prices, data_prefix="../../data/stockx/"):

return sx_prices

def annotate_score(margins, score_mode):
for item in margins:
if score_mode == 'naive':
score, volume = score_crossing_margin_rate(margins[item])
elif score_mode == 'multi':
score, volume = score_margin_single_entity_transactions_size(margins[item])
margins[item]['score'] = score
margins[item]['volume'] = volume
return margins

def score_crossing_margin_rate(item):
return item['crossing_margin_rate'], 0

def score_margin_single_entity_transactions_size(item):
# an item with a high crossing margin rate, how single entity price, high
# transaction amounts and closer to norm size scores better

# we can't really accurately account for transaction rate as some data is
# still being scraped. when we don't have such data we use the this table
# as approximation

size_discount_multiplier = {
'3.5': 0.40,
'4': 0.50,
'4.5': 0.60,
'5': 0.70,
'5.5': 0.75,
'6': 0.80,
'6.5': 0.85,
'7': 0.90,
'7.5': 0.95,
'8': 0.98,
'8.5': 1.00,
'9': 1.00,
'9.5': 1.00,
'10': 1.00,
'10.5': 1.00,
'11': 1.00,
'11.5': 0.98,
'12': 0.95,
'12.5': 0.90,
'13': 0.85,
'13.5': 0.80,
'14': 0.75,
'14.5': 0.70,
'15': 0.70,
'16': 0.60,
'17': 0.50,
'18': 0.40
}

if len(item['sx_transactions']) > 0:
duration = datetime.datetime.now(
pytz.timezone('America/New_York')) - datetime.datetime.strptime(
item['sx_transactions'][-1]['time'], "%Y-%m-%dT%H:%M:%S%z")
transaction_rate = len(item['sx_transactions']) * 3600 * 24 // duration.total_seconds()
else:
if item['shoe_size'] in size_discount_multiplier:
transaction_rate = 0.2 * size_discount_multiplier[item['shoe_size']]
else:
print('unknown size {}'.format(item['shoe_size']))
transaction_rate = 0

price_discount = 1
if item['sx_best_ask'] > 1000:
price_discount = 0.3
elif item['sx_best_ask'] > 500:
price_discount = 0.6
elif item['sx_best_ask'] > 300:
price_discount = 0.9

return item['crossing_margin_rate'] * transaction_rate * price_discount, transaction_rate

def generate_html_report(score_sorted_item, limit, **run_info):
text = ("<html><head></head><body>"
"Hi,<br><p>Please see below for a list of candidate shoes.</p>")

text += ("<table>"
"<tr>"
"<th>Name</th>"
"<th>Size</th>"
"<th>StockX</th>"
"<th>FlightClub</th>"
"<th>Crossing Margin</th>"
"<th>Crossing Margin Rate</th>"
"<th>StockX Transaction Rate</th>"
"<th>Score</th>"
"<th>Style ID</th>"
"</tr>")

report_cnt = 0
for key in score_sorted_item:
shoe = key[1]
if shoe["crossing_margin_rate"] > 0 and (not limit or report_cnt < limit):
text += ("<tr>"
"<td>{}</td>"
"<td>{}</td>"
"<td>{}</td>"
"<td>{}</td>"
"<td>{:.2f}</td>"
"<td>{:.2f}</td>"
"<td>{:.2f}</td>"
"<td>{:.2f}</td>"
"<td>{}</td>"
"</tr>").format(shoe['name'], shoe['shoe_size'],
"<a href=\"{}\">{:.2f}</a>".format(
shoe['sx_url'], float(shoe['sx_best_ask'])),
"<a href=\"{}\">{:.2f}</a>".format(
shoe['fc_url'], float(shoe['fc_px'])),
shoe['crossing_margin'], shoe['crossing_margin_rate'],
shoe['volume'], shoe['score'], shoe['style_id'])
report_cnt += 1

text += "<table><br><br>{}".format(pp.pformat(run_info))
text += "<br><br>Thanks,<br>Sneaky Bot</body></html>"

return text if report_cnt > 0 else "But didn't find anything :("

if __name__ == "__main__":
parser = argparse.ArgumentParser()
fc_file = "flightclub.20190614.txt"
sx_file = "best_prices.20190614-230824.txt"
sx_transactions_folder = "../../data/stockx/20190615-145954/"

parser.add_argument(
"--score_mode",
help=("[multi|naive] with what criteria we decide how much we want to "
" buy this pair shoes"))
parser.add_argument(
"--emails",
help="comma separated list of email addresses to send report")
parser.add_argument(
"--limit",
help="max number of items to send in one email")
parser.add_argument(
"--out",
help="output file to write to")
args = parser.parse_args()

runtime = datetime.datetime.now().strftime("%Y%m%d-%H%M%S")

fc_prices, sx_prices = read_files("flightclub.20190614.txt", "best_prices.txt")
sx_prices = annotate_transaction_history(sx_prices, "../../data/stockx/20190615-145954/")
fc_prices, sx_prices = read_files(fc_file, sx_file)
sx_prices = annotate_transaction_history(sx_prices, sx_transactions_folder)
matches = match_items(fc_prices, sx_prices)

margins = find_margin(matches)
score_mode = args.score_mode if args.score_mode else 'multi'
margins = annotate_score(margins, score_mode)

score_sorted_item = sorted(
margins.items(), key=lambda kv: kv[1]['score'], reverse=True)
outfile = args.out if args.out else "score_sorted.{}.txt".format(runtime)
with open(outfile, "w") as wfile:
wfile.write(pp.pformat(score_sorted_item))

if args.emails:
report = generate_html_report(
score_sorted_item, args.limit, runtime=runtime, score=score_mode,
fc_file=fc_file, sx_file=sx_file,
sx_transactions_folder=sx_transactions_folder)

server = smtplib.SMTP('smtp.gmail.com:587')
server.ehlo()
server.starttls()
server.login('testname.zhehao@gmail.com', 'test@2019')

msg = MIMEText(report, 'html')
msg['Subject'] = 'Check out these shoes %s' % runtime
msg['From'] = 'testname.zhehao@gmail.com'
msg['To'] = args.emails

# Send the message via our own SMTP server.
server.send_message(msg)
server.quit()

# "regular sized" shoes usually have low profit margin. we should revisit
# this from time to time

# regular_size_crossing_margin_rate_sorted_items = []
# for item in score_sorted_item:
# shoe_size = float(item[1]['shoe_size'])
# if shoe_size == 9.0 or shoe_size == 8.5 or shoe_size == 9.5:
# regular_size_crossing_margin_rate_sorted_items.append(item[1])

# with open("regular_size_crossing_margin_sorted.{}.txt".format(runtime), "w") as wfile:
# wfile.write(pp.pformat(regular_size_crossing_margin_rate_sorted_items))

crossing_margin_rate_sorted_items = sorted(margins.items(), key=lambda kv: kv[1]['crossing_margin_rate'], reverse=True)
with open("crossing_margin_sorted.{}.txt".format(runtime), "w") as wfile:
wfile.write(pp.pformat(crossing_margin_rate_sorted_items))

regular_size_crossing_margin_rate_sorted_items = []
for item in crossing_margin_rate_sorted_items:
shoe_size = float(item[1]['shoe_size'])
if shoe_size == 9.0 or shoe_size == 8.5 or shoe_size == 9.5:
regular_size_crossing_margin_rate_sorted_items.append(item[1])

with open("regular_size_crossing_margin_sorted.{}.txt".format(runtime), "w") as wfile:
wfile.write(pp.pformat(regular_size_crossing_margin_rate_sorted_items))

10 changes: 7 additions & 3 deletions src/stockx/stockx_feed.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,10 @@
def sanitize_key(key):
return key.lower().replace('-', '').replace(' ', '')

def sanitize_filename(key):
return key.replace(
'(', '-').replace(')', '-').replace(' ', '-').replace('/', '-')

class StockXFeed():
def __init__(self):
self.stockx = Stockx()
Expand Down Expand Up @@ -297,10 +301,10 @@ def find_promising(items_map):
transactions_json)

book_filename = os.path.join(
outdir, style_id + '.txt')
outdir, sanitize_filename(style_id + '.txt'))
transactions_filename = os.path.join(
outdir, (style_id + '.transaction.txt').replace(
'(', '-').replace(')', '-'))
outdir, sanitize_filename(
style_id + '.transaction.txt'))

with open(book_filename, 'w') as wfile, open(transactions_filename, 'w') as tfile:
wfile.write(bookstr)
Expand Down

0 comments on commit 6686f49

Please sign in to comment.