def test_behaviour_may_change_with_batch_or_stream_processing_using_euclidean_distance_1(
self):
data = [(item.get('x'), item.get('y')) for item in self.sample_1()]
stream_result = pip(data, 300, distance='euclidean', stream_mode=True)
batch_result = pip(data, 300, distance='euclidean', stream_mode=False)
self.assertNotAlmostEqualCurves(stream_result, batch_result)
def test_behaviour_may_change_with_batch_or_stream_processing_using_euclidean_distance_3(
self):
data = self.sample_3()
stream_result = pip(data, 300, distance='euclidean', stream_mode=True)
batch_result = pip(data, 300, distance='euclidean', stream_mode=False)
self.assertNotAlmostEqualCurves(stream_result, batch_result)
def test_result_should_always_be_the_same_for_a_given_input(self):
data = [(item.get('x'), item.get('y')) for item in self.sample_1()]
data_length = len(data)
first_result = pip(data, data_length/10)
for i in range(3):
self.assertAlmostEqualCurves(
self.pip(data, data_length/10), first_result, "Result has changed on iteration {}".format(i+1)
)
def test_result_should_always_be_the_same_for_a_given_input(self):
data = [(item.get('x'), item.get('y')) for item in self.sample_1()]
data_length = len(data)
first_result = pip(data, data_length / 10)
for i in range(3):
self.assertAlmostEqualCurves(
self.pip(data, data_length / 10), first_result,
"Result has changed on iteration {}".format(i + 1))
def patterntype(x):
if x == 1: # Double Bottom Pattern
return pip(__PATTERN_DOUBLE_BOTTOM, 7), 7
if x == 2: #
return pip(__CUP_WITH_HANDLE, 7), 7
if x == 3: # Scallops Pattern
return pip(__PATTERN_SCALLOPS, 7), 7
if x == 4: # Three Rising Valleys Pattern
return pip(__THREE_RISING_VALLEYS, 7), 7
if x == 5: # Round Bottoms Pattern
return pip(__ROUNDING_BOTTOMS, 7), 7
if x == 6: # Triangles Descending Pattern
return pip(__TRIANGLES_DESCENDING, 7), 7
if x == 7: #
return pip(__ROUNDING_TOPS, 7), 7
if x == 8: #
return pip(__BUMP_AND_RUN_REVERSAL_BOTTOM, 7), 7
if x == 9: # Triple Bottom Pattern
return pip(__PATTERN_TRIPLE_BOTTOM, 10), 10
else: # DEFAULT. exception route
return None
def findpattern(p_type, p_low, p_high, period):
# p_type : pattern type index
# p_low : low price; it's used to search shcode
# p_high : high price; it's used to search shcode
# period : pattern period
start_time = time.time()
compare, n = pttype.patterntype(p_type) # choose pattern by p_type
rows = shquery.makequerylist(p_low, p_high) # stock filtering
print("p_low=", p_low, "와 p_high=", p_high, " 조건에 부합한 종목코드 : ", rows)
data_array = []
for row in rows: # each stock
tmps = shquery.getstockprice(row) # get stock data from DB
temp_array = [] # stock price data
pip_array = [] # list of found index
sim_array = [] # list of found index
i = -1
for tmp in tmps: # append data for PIP
i += 1
temp_array.append([i, int(tmp['closeprice'])])
for k in range(0, len(temp_array) - period): # find relevant index
result = pip(temp_array[k:k + period], n)
r_row, p_value = pearsonr(
numpy.array(result)[:, 1],
numpy.array(compare)[:, 1])
if r_row > 0.7:
print("종목코드 ", row, "에서 index가 ", k, "일 때, 피어슨상관계수 : ", r_row)
pip_array.append(k) # k is index
sim_array.append(r_row)
# fill data
if pip_array:
data_array.append(row)
data_array.append(pip_array)
data_array.append(sim_array)
print("┌기간=", period, "에서 패턴발견 알고리즘 적용결과 리스트 구현")
for i in range(0, len(data_array)):
print(data_array[i])
# measure of running-time
print("--- processing time : %s seconds ---" % (time.time() - start_time))
return data_array
def test_behaviour_changes_according_to_the_distance_function_used_2(self):
data = self.sample_2()
vertical_result = pip(data, 300, distance='vertical')
euclidean_result = pip(data, 300, distance='euclidean')
self.assertNotAlmostEqualCurves(vertical_result, euclidean_result)
def test_behaviour_changes_according_to_the_distance_function_used_1(self):
data = [(item.get('x'), item.get('y')) for item in self.sample_1()]
vertical_result = pip(data, 300, distance='vertical')
euclidean_result = pip(data, 300, distance='euclidean')
self.assertNotAlmostEqualCurves(vertical_result, euclidean_result)
def test_behaviour_may_change_with_batch_or_stream_processing_using_euclidean_distance_3(self):
data = self.sample_3()
stream_result = pip(data, 300, distance='euclidean', stream_mode=True)
batch_result = pip(data, 300, distance='euclidean', stream_mode=False)
self.assertNotAlmostEqualCurves(stream_result, batch_result)
def test_behaviour_may_change_with_batch_or_stream_processing_using_euclidean_distance_1(self):
data = [(item.get('x'), item.get('y')) for item in self.sample_1()]
stream_result = pip(data, 300, distance='euclidean', stream_mode=True)
batch_result = pip(data, 300, distance='euclidean', stream_mode=False)
self.assertNotAlmostEqualCurves(stream_result, batch_result)
def get_pip_dates(df, k):
return _pip_array_to_dates(df, pip(_pandas_to_xy_array(df), k))
def test_behaviour_changes_according_to_the_distance_function_used_1(self):
data = [(item.get('x'), item.get('y')) for item in self.sample_1()]
vertical_result = pip(data, 300, distance='vertical')
euclidean_result = pip(data, 300, distance='euclidean')
self.assertNotAlmostEqualCurves(vertical_result, euclidean_result)
def test_behaviour_changes_according_to_the_distance_function_used_3(self):
data = self.sample_3()
vertical_result = pip(data, 300, distance='vertical')
euclidean_result = pip(data, 300, distance='euclidean')
self.assertNotAlmostEqualCurves(vertical_result, euclidean_result)
# Fetch
rows = cursor.fetchall()
i = -1
array = []
graph = []
for row in rows:
#print(row)
i = i + 1
#print(pip([i, int(row['closeprice'])], 5))
#print([i, int(row['closeprice'])])
array.append([i, int(row['closeprice'])])
graph.append(int(row['closeprice']))
#print(array)
# Perceptually Important Points
result = pip(array, 7)
print(result)
plt.plot(graph)
# Perceptually Important Points
index = []
data = []
for idx in result:
i = i + 1
index.append(idx[0])
data.append(idx[1])
#index = [0,43,101,174,339,456,499]
#data = [5500, 6470, 9870, 7080, 10800, 7280, 7950]
plt.plot(index, data)
plt.grid(True)
plt.show()
graph_array = []
sample_array = [[0, 8000], [1, 4000], [2, 4000], [3, 8000], [4, 4000],
[5, 4000], [6, 8000]] # Double Bottom graph
#sample_array = [[0, 8000], [1, 5500], [2, 6500], [3, 5500], [4, 6000], [5, 5000], [6, 4000]] # Triangles,Descending, up breakout (6)
#sample_array = [[0, 6000], [1, 5500], [2, 6500], [3, 6000], [4, 7000], [5, 6500], [6, 8000]] # Three Rising Valleys, up breakout (4)
#sample_array = [[0, 8000], [1, 6000], [2, 5500], [3, 5000], [4, 5500], [5, 6000], [6, 8000]] # Rounding Bottoms, up breakout (5) 그림이 좀 이상...
#sample_array = [[0, 4000], [1, 5250], [2, 6500], [3, 5000], [4,6500], [5, 6000], [6, 7500]] # Cup with Handle
#sample_array = [[0, 4000], [1, 4700], [2, 5400], [3, 6000], [4,5400], [5, 4700], [6, 7000]] # Rounding Tops
#sample_array = [[0, 6000], [1, 6500], [2, 6000], [3, 6250], [4,5000], [5, 6250], [6, 6000]] # Bump-and-Run Reversal Bottoms
for row in rows:
i += 1
graph_array.append([i, int(row['closeprice'])]) # append data for PIP
# Find Pattern using pearson correlation coefficient
sample = pip(sample_array, 7)
plt.plot(numpy.array(sample)[:, 0],
numpy.array(sample)[:, 1]) # plot result of sample pip
# plt.show()
pip_array = [] # list of found index
for k in range(0, len(graph_array) - 30):
result = pip(graph_array[k:k + 30], 7)
a = numpy.array(result) # print(a[:, 1])
r_row, p_value = pearsonr(
numpy.array(result)[:, 1],
numpy.array(sample)[:, 1])
if r_row > 0.7:
print("종목코드 ", shcode, "에서 index가 ", k, "일 때, 피어슨상관계수 : ", r_row)
print(rows[k])
pip_array.append(k)
import pymysql
from fastpip import pip
import mysqlinfo
# Open database connection
myhost, myport, myuser, mypassword, mydb, mycharset = mysqlinfo.getmysqlinfo()
conn = pymysql.connect(host=myhost, port=myport, user=myuser, password=mypassword, db=mydb, charset=mycharset)
# prepare a cursor object using cursor() method
cursor = conn.cursor()
# execute SQL query using execute() method.
sql = "SELECT * FROM stockprice WHERE shcode='000020'"
cursor.execute(sql)
# Fetch a single row using fetchone() method.
rows = cursor.fetchall()
# print(rows) # total rows
# print(rows[0]) # first row
# print(rows[1]) # second row
i = -1
for row in rows:
print(row)
i = i + 1
print(pip([i,int(row['closeprice'])], 5))
# disconnect from server
conn.close()