def max_profits(self,symbol="AAPL",from_date="2014-01-01", to_date="2015-01-01"):
"""
the max profits with unlimited transactions in date range [from_date, to_date)
only calculate open price for simplification
"""
rows = db_conn.execute_select_sql_file('./service_sqls/open_prices.sql',symbol,from_date,to_date)
prices = [rows[j][1] for j in xrange(len(rows))]
date_format = "{:%Y-%m-%d}"
records = []
max_p = 0
i = 0
while(i < len(prices)):
while(i+1< len(prices) and prices[i] >= prices[i+1]):
i += 1
low = prices[i]
buy_day = rows[i][2]
while(i+1 < len(prices) and prices[i]<=prices[i+1]):
i += 1
high = prices[i]
sell_day = rows[i][2]
p = high - low
if(p > 0):
records.append((date_format.format(buy_day),date_format.format(sell_day),p))
max_p += p
i += 1
return {"symbol":symbol,"from_date":from_date,"to_date":to_date,"max_profits":max_p,"records":records}
def top_industries(self,change_percentiles=0.3,count=3,from_date = "2014-01-01", to_date="2014-07-01"):
"""
the top {count} sectores whose price changed over {change_percentiles} in date range [from_date, to_date)
return ['symbol1', 'symbol2']
date format "YYYY-MM-DD"
"""
# top industries which has the most stocks whose price has change over {change_percentiles}
rows = db_conn.execute_select_sql_file('./service_sqls/top_industries.sql',change_percentiles,from_date,to_date)
if count >= len(rows):
return [rows[i][0] for i in xrange(len(rows))]
else:
return [rows[i][0] for i in xrange(count)]
def top_change_stocks(self,count=10, from_date="2014-07-01", to_date="2014-10-01"):
"""
the top {count} stocks whose price changes the most in date range [from_date, to_date)
date format "YYYY-MM-DD"
return ['symbol1', 'symbol2',...]
"""
# sort by the varience of each stock's 5 prices
rows = db_conn.execute_select_sql_file('./service_sqls/stock_varience.sql',from_date,to_date)
if count >= len(rows):
return [rows[i][0] for i in xrange(len(rows))]
else:
return [rows[i][0] for i in xrange(count)]
def max_profits_k_transaction(self,k = 2,symbol="AAPL",from_date="2014-01-01", to_date="2015-01-01"):
"""
the max profits by maximum k transactions in date range [from_date, to_date)
only calculate open price for simplification
"""
rows = db_conn.execute_select_sql_file('./service_sqls/open_prices.sql',symbol,from_date,to_date)
prices = [rows[j][1] for j in xrange(len(rows))]
date_format = "{:%Y-%m-%d}"
n = len(prices)
if n<=1:
return {"symbol":symbol,"from_date":from_date,"to_date":to_date,"max_profits":0,"records":[], "transaction_times":k}
if (k >= n / 2):
quick_res = self.quick_solve(prices)
return {"symbol":symbol,"from_date":from_date,"to_date":to_date,"max_profits":quick_res["max_profit"],"records":quick_res["records"], "transaction_times":k}
dp = [[0 for j in xrange(n)] for i in xrange(k+1)]
# -1 means not in the transaction, 0 means to hold, 1 means to sell
last_records = [[-1 for _ in xrange(n)] for _ in xrange(n)]
cur_records = [[-1 for _ in xrange(n)] for _ in xrange(n)]
cur_record = [-1 for _ in xrange(n)]
for i in range(1,k+1):
hold_idx = 0
temp = dp[i-1][0] - prices[0]
for j in xrange(1,n):
# equals to j-1 th transaction or sell to complete new transaction.
if(temp+prices[j] > dp[i][j-1]):
dp[i][j] = temp+prices[j]
# new transaction complete
cur_record[j] = 1
# clear records in range [hold_idx, j)
self.clear_record(cur_record,hold_idx, j)
# should copy the last_records[hold_idx]
self.copy_record(last_records[hold_idx-1], cur_record, 0, hold_idx)
else:
dp[i][j] = dp[i][j-1]
# update cur_records to save cur_record
self.copy_record(cur_record,cur_records[j],0,n)
# potential hold should cost least, which means to remain most profit after hold.
if(dp[i-1][j]-prices[j] > temp):
temp = dp[i-1][j] - prices[j]
hold_idx = j
last_records, cur_records = cur_records, last_records
return {"symbol":symbol,"from_date":from_date,"to_date":to_date,"max_profits":dp[k][n-1],"records":cur_record, "transaction_times":k}
def top_change_stocks(self,
count=10,
from_date="2014-07-01",
to_date="2014-10-01"):
"""
the top {count} stocks whose price changes the most in date range [from_date, to_date)
date format "YYYY-MM-DD"
return ['symbol1', 'symbol2',...]
"""
# sort by the varience of each stock's 5 prices
rows = db_conn.execute_select_sql_file(
'./service_sqls/stock_varience.sql', from_date, to_date)
if count >= len(rows):
return [rows[i][0] for i in xrange(len(rows))]
else:
return [rows[i][0] for i in xrange(count)]
def top_industries(self,
change_percentiles=0.3,
count=3,
from_date="2014-01-01",
to_date="2014-07-01"):
"""
the top {count} sectores whose price changed over {change_percentiles} in date range [from_date, to_date)
return ['symbol1', 'symbol2']
date format "YYYY-MM-DD"
"""
# top industries which has the most stocks whose price has change over {change_percentiles}
rows = db_conn.execute_select_sql_file(
'./service_sqls/top_industries.sql', change_percentiles, from_date,
to_date)
if count >= len(rows):
return [rows[i][0] for i in xrange(len(rows))]
else:
return [rows[i][0] for i in xrange(count)]
def max_profits(self,
symbol="AAPL",
from_date="2014-01-01",
to_date="2015-01-01"):
"""
the max profits with unlimited transactions in date range [from_date, to_date)
only calculate open price for simplification
"""
rows = db_conn.execute_select_sql_file(
'./service_sqls/open_prices.sql', symbol, from_date, to_date)
prices = [rows[j][1] for j in xrange(len(rows))]
date_format = "{:%Y-%m-%d}"
records = []
max_p = 0
i = 0
while (i < len(prices)):
while (i + 1 < len(prices) and prices[i] >= prices[i + 1]):
i += 1
low = prices[i]
buy_day = rows[i][2]
while (i + 1 < len(prices) and prices[i] <= prices[i + 1]):
i += 1
high = prices[i]
sell_day = rows[i][2]
p = high - low
if (p > 0):
records.append((date_format.format(buy_day),
date_format.format(sell_day), p))
max_p += p
i += 1
return {
"symbol": symbol,
"from_date": from_date,
"to_date": to_date,
"max_profits": max_p,
"records": records
}
def max_profits_k_transaction(self,
k=2,
symbol="AAPL",
from_date="2014-01-01",
to_date="2015-01-01"):
"""
the max profits by maximum k transactions in date range [from_date, to_date)
only calculate open price for simplification
"""
rows = db_conn.execute_select_sql_file(
'./service_sqls/open_prices.sql', symbol, from_date, to_date)
prices = [rows[j][1] for j in xrange(len(rows))]
date_format = "{:%Y-%m-%d}"
n = len(prices)
if n <= 1:
return {
"symbol": symbol,
"from_date": from_date,
"to_date": to_date,
"max_profits": 0,
"records": [],
"transaction_times": k
}
if (k >= n / 2):
quick_res = self.quick_solve(prices)
return {
"symbol": symbol,
"from_date": from_date,
"to_date": to_date,
"max_profits": quick_res["max_profit"],
"records": quick_res["records"],
"transaction_times": k
}
dp = [[0 for j in xrange(n)] for i in xrange(k + 1)]
# -1 means not in the transaction, 0 means to hold, 1 means to sell
last_records = [[-1 for _ in xrange(n)] for _ in xrange(n)]
cur_records = [[-1 for _ in xrange(n)] for _ in xrange(n)]
cur_record = [-1 for _ in xrange(n)]
for i in range(1, k + 1):
hold_idx = 0
temp = dp[i - 1][0] - prices[0]
for j in xrange(1, n):
# equals to j-1 th transaction or sell to complete new transaction.
if (temp + prices[j] > dp[i][j - 1]):
dp[i][j] = temp + prices[j]
# new transaction complete
cur_record[j] = 1
# clear records in range [hold_idx, j)
self.clear_record(cur_record, hold_idx, j)
# should copy the last_records[hold_idx]
self.copy_record(last_records[hold_idx - 1], cur_record, 0,
hold_idx)
else:
dp[i][j] = dp[i][j - 1]
# update cur_records to save cur_record
self.copy_record(cur_record, cur_records[j], 0, n)
# potential hold should cost least, which means to remain most profit after hold.
if (dp[i - 1][j] - prices[j] > temp):
temp = dp[i - 1][j] - prices[j]
hold_idx = j
last_records, cur_records = cur_records, last_records
return {
"symbol": symbol,
"from_date": from_date,
"to_date": to_date,
"max_profits": dp[k][n - 1],
"records": cur_record,
"transaction_times": k
}
