# Local Periodic Patterns, which are patterns (sets of events) that have a periodic behavior in some non predefined
# time-intervals. A pattern is said to be a local periodic pattern if it appears regularly and continuously in some
# time-intervals. The maxSoPer (maximal period of spillovers) measure allows detecting time-intervals of variable
# lengths where a pattern is continuously periodic, while the minDur (minimal duration) measure ensures that those
# time-intervals have a minimum duration.
#
# **Importing this algorithm into a python program**
# --------------------------------------------------------
#
#
# from PAMI.localPeriodicPattern.basic import LPPMDepth as alg
#
# obj = alg.LPPMDepth(iFile, maxPer, maxSoPer, minDur)
#
# obj.mine()
#
# localPeriodicPatterns = obj.getPatterns()
#
# print(f'Total number of local periodic patterns: {len(localPeriodicPatterns)}')
#
# obj.save(oFile)
#
# Df = obj.getPatternsAsDataFrame()
#
# memUSS = obj.getMemoryUSS()
#
# print(f'Total memory in USS: {memUSS}')
#
# memRSS = obj.getMemoryRSS()
#
# print(f'Total memory in RSS: {memRSS}')
#
# runtime = obj.getRuntime()
#
# print(f'Total execution time in seconds: {runtime})
#
__copyright__ = """
Copyright (C) 2021 Rage Uday Kiran
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
Copyright (C) 2021 Rage Uday Kiran
"""
from PAMI.localPeriodicPattern.basic import abstract as _ab
from typing import List, Dict, Tuple, Set, Union, Any, Generator
import pandas as pd
from deprecated import deprecated
[docs]
class LPPMDepth(_ab._localPeriodicPatterns):
"""
:Description:
Local Periodic Patterns, which are patterns (sets of events) that have a periodic behavior in some non predefined
time-intervals. A pattern is said to be a local periodic pattern if it appears regularly and continuously in some
time-intervals. The maxSoPer (maximal period of spillovers) measure allows detecting time-intervals of variable
lengths where a pattern is continuously periodic, while the minDur (minimal duration) measure ensures that those
time-intervals have a minimum duration.
:Reference:
Fournier-Viger, P., Yang, P., Kiran, R. U., Ventura, S., Luna, J. M.. (2020). Mining Local Periodic Patterns in
a Discrete Sequence. Information Sciences, Elsevier, to appear. [ppt] DOI: 10.1016/j.ins.2020.09.044
:param iFile: str :
Name of the Input file to mine complete set of local periodic pattern's
:param oFile: str :
Name of the output file to store complete set of local periodic patterns
:param minDur: str:
Minimal duration in seconds between consecutive periods of time-intervals where a pattern is continuously periodic.
:param maxPer: float:
Controls the maximum number of transactions in which any two items within a pattern can reappear.
:param maxSoPer: float:
Controls the maximum number of time periods between consecutive periods of time-intervals where a pattern is continuously periodic.
:param sep: str :
This variable is used to distinguish items from one another in a transaction. The default seperator is tab space. However, the users can override their default separator.
:Attributes:
iFile : str
Input file name or path of the input file
oFile : str
Output file name or path of the output file
maxPer : float
User defined maxPer value.
maxSoPer : float
User defined maxSoPer value.
minDur : float
User defined minDur value.
tsmin : int / date
First time stamp of input data.
tsmax : int / date
Last time stamp of input data.
startTime : float
Time when start of execution the algorithm.
endTime : float
Time when end of execution the algorithm.
finalPatterns : dict
To store local periodic patterns and its PTL.
tsList : dict
To store items and its time stamp as bit vector.
sep : str
separator used to distinguish items from each other. The default separator is tab space.
:Methods:
createTSlist()
Create the TSlist as bit vector from input data.
generateLPP()
Generate 1 length local periodic pattens by TSlist and execute depth first search.
calculatePTL(tsList)
Calculate PTL from input tsList as bit vector
LPPMDepthSearch(extensionOfP)
Mining local periodic patterns using depth first search.
mine()
Mining process will start from here.
getMemoryUSS()
Total amount of USS memory consumed by the mining process will be retrieved from this function.
getMemoryRSS()
Total amount of RSS memory consumed by the mining process will be retrieved from this function.
getRuntime()
Total amount of runtime taken by the mining process will be retrieved from this function.
getLocalPeriodicPatterns()
return local periodic patterns and its PTL
save(oFile)
Complete set of local periodic patterns will be loaded in to an output file.
getPatternsAsDataFrame()
Complete set of local periodic patterns will be loaded in to a dataframe.
**Executing the code on terminal:**
--------------------------------------
.. code-block:: console
Format:
(.venv) $ python3 LPPMDepth.py <inputFile> <outputFile> <maxPer> <minSoPer> <minDur>
Example Usage:
(.venv) $ python3 LPPMDepth.py sampleDB.txt patterns.txt 0.3 0.4 0.5
.. note: minDur will be considered as time interval between two consecutive periods
**Sample run of importing the code:**
----------------------------------------
.. code-block:: python
from PAMI.localPeriodicPattern.basic import LPPMDepth as alg
obj = alg.LPPMDepth(iFile, maxPer, maxSoPer, minDur)
obj.mine()
localPeriodicPatterns = obj.getPatterns()
print(f'Total number of local periodic patterns: {len(localPeriodicPatterns)}')
obj.save(oFile)
Df = obj.getPatternsAsDataFrame()
memUSS = obj.getMemoryUSS()
print(f'Total memory in USS: {memUSS}')
memRSS = obj.getMemoryRSS()
print(f'Total memory in RSS: {memRSS}')
runtime = obj.getRuntime()
print(f'Total execution time in seconds: {runtime})
**Credits:**
-------------
The complete program was written by So Nakamura under the supervision of Professor Rage Uday Kiran.
"""
_localPeriodicPatterns__iFile = ''
_localPeriodicPatterns__oFile = ''
_localPeriodicPatterns__maxPer = str()
_localPeriodicPatterns__maxSoPer = str()
_localPeriodicPatterns__minDur = str()
__tsmin = 0
__tsmax = 0
_localPeriodicPatterns__startTime = float()
_localPeriodicPatterns__endTime = float()
_localPeriodicPatterns__memoryUSS = float()
_localPeriodicPatterns__memoryRSS = float()
_localPeriodicPatterns__finalPatterns = {}
__tsList = {}
_localPeriodicPatterns__sep = ' '
__Database = []
def __creatingItemSets(self) -> None:
"""
Storing the complete transactions of the database/input file in a database variable
"""
self.__Database = []
if isinstance(self._localPeriodicPatterns__iFile, _ab._pd.DataFrame):
if self._localPeriodicPatterns__iFile.empty:
print("its empty..")
i = self._localPeriodicPatterns__iFile.columns.values.tolist()
if 'Transactions' in i:
self.__Database = self._localPeriodicPatterns__iFile['Transactions'].tolist()
if 'Patterns' in i:
self.__Database = self._localPeriodicPatterns__iFile['Patterns'].tolist()
if isinstance(self._localPeriodicPatterns__iFile, str):
if _ab._validators.url(self._localPeriodicPatterns__iFile):
data = _ab._urlopen(self._localPeriodicPatterns__iFile)
for line in data:
line.strip()
line = line.decode("utf-8")
temp = [i.rstrip() for i in line.split(self._localPeriodicPatterns__sep)]
temp = [x for x in temp if x]
self.__Database.append(temp)
else:
try:
with open(self._localPeriodicPatterns__iFile, 'r', encoding='utf-8') as f:
for line in f:
line.strip()
temp = [i.rstrip() for i in line.split(self._localPeriodicPatterns__sep)]
temp = [x for x in temp if x]
self.__Database.append(temp)
except IOError:
print("File Not Found")
quit()
def __createTSlist(self) -> None:
"""
Create tsList as bit vector from temporal data.
"""
count = 1
bitVector = None
for line in self.__Database:
bitVector = 0b1 << count
bitVector = bitVector | 0b1
ts = line[0]
for item in line[1:]:
if self.__tsList.get(item):
different = abs(bitVector.bit_length() - self.__tsList[item].bit_length())
self.__tsList[item] = self.__tsList[item] << different
self.__tsList[item] = self.__tsList[item] | 0b1
else:
self.__tsList[item] = bitVector
count += 1
self.__tsmax = int(ts)
for item in self.__tsList:
different = abs(bitVector.bit_length() - self.__tsList[item].bit_length())
self.__tsList[item] = self.__tsList[item] << different
self._localPeriodicPatterns__maxPer = self.__convert(self._localPeriodicPatterns__maxPer)
self._localPeriodicPatterns__maxSoPer = self.__convert(self._localPeriodicPatterns__maxSoPer)
self._localPeriodicPatterns__minDur = self.__convert(self._localPeriodicPatterns__minDur)
def __generateLPP(self) -> None:
"""
Generate local periodic items from bit vector tsList.
When finish generating local periodic items, execute mining depth first search.
"""
I = set()
PTL = {}
for item in self.__tsList:
PTL[item] = set()
ts = list(bin(self.__tsList[item]))
ts = ts[2:]
start = -1
currentTs = 1
tsPre = ' '
soPer = ' '
for t in ts[currentTs:]:
if t == '0':
currentTs += 1
continue
else:
tsPre = currentTs
currentTs += 1
break
for t in ts[currentTs:]:
if t == '0':
currentTs += 1
continue
else:
per = currentTs - tsPre
if per <= self._localPeriodicPatterns__maxPer and start == -1:
start = tsPre
soPer = self._localPeriodicPatterns__maxSoPer
if start != -1:
soPer = max(0, soPer + per - self._localPeriodicPatterns__maxPer)
if soPer > self._localPeriodicPatterns__maxSoPer:
if tsPre - start >= self._localPeriodicPatterns__minDur:
PTL[item].add((start, tsPre))
"""else:
bitVector = 0b1 << currentTs
different = abs(self.tsList[item].bit_length() - bitVector.bit_length())
bitVector = bitVector | 0b1
bitVector = bitVector << different
self.tsList[item] = self.tsList[item] | bitVector"""
start = -1
tsPre = currentTs
currentTs += 1
if start != -1:
soPer = max(0, soPer + self.__tsmax - tsPre - self._localPeriodicPatterns__maxPer)
if soPer > self._localPeriodicPatterns__maxSoPer and tsPre - start >= self._localPeriodicPatterns__minDur:
PTL[item].add((start, tsPre))
"""else:
bitVector = 0b1 << currentTs+1
different = abs(self.tsList[item].bit_length() - bitVector.bit_length())
bitVector = bitVector | 0b1
bitVector = bitVector << different
self.tsList[item] = self.tsList[item] | bitVector"""
if soPer <= self._localPeriodicPatterns__maxSoPer and self.__tsmax - start >= self._localPeriodicPatterns__minDur:
PTL[item].add((start, self.__tsmax))
"""else:
bitVector = 0b1 << currentTs+1
different = abs(self.tsList[item].bit_length() - bitVector.bit_length())
bitVector = bitVector | 0b1
bitVector = bitVector << different
self.tsList[item] = self.tsList[item] | bitVector"""
if len(PTL[item]) > 0:
I |= {item}
self._localPeriodicPatterns__finalPatterns[item] = PTL[item]
I = sorted(list(I))
# I = set(I)
self.__LPPMDepthSearch(I)
def __calculatePTL(self, tsList: int) -> Set[Tuple[int, int]]:
"""
calculate PTL from tsList as bit vector.
:param tsList: it is one item's tsList which is used bit vector.
:type tsList: int
:return: it is PTL of input item.
:rtype: set
"""
tsList = list(bin(tsList))
tsList = tsList[2:]
start = -1
currentTs = 1
PTL = set()
tsPre = ' '
soPer = ' '
for ts in tsList[currentTs:]:
if ts == '0':
currentTs += 1
continue
else:
tsPre = currentTs
currentTs += 1
break
for ts in tsList[currentTs:]:
if ts == '0':
currentTs += 1
continue
else:
per = currentTs - tsPre
if per <= self._localPeriodicPatterns__maxPer and start == -1:
start = tsPre
soPer = self._localPeriodicPatterns__maxSoPer
if start != -1:
soPer = max(0, soPer + per - self._localPeriodicPatterns__maxPer)
if soPer > self._localPeriodicPatterns__maxSoPer:
if tsPre - start >= self._localPeriodicPatterns__minDur:
PTL.add((start, tsPre))
start = -1
tsPre = currentTs
currentTs += 1
if start != -1:
soPer = max(0, soPer + self.__tsmax - tsPre - self._localPeriodicPatterns__maxPer)
if soPer > self._localPeriodicPatterns__maxSoPer and tsPre - start >= self._localPeriodicPatterns__minDur:
PTL.add((start, tsPre))
if soPer <= self._localPeriodicPatterns__maxSoPer and self.__tsmax - start >= self._localPeriodicPatterns__minDur:
PTL.add((start, tsPre))
return PTL
def __LPPMDepthSearch(self, extensionsOfP: List[Union[Tuple[str, ...], str]]) -> None:
"""
Mining n-length local periodic pattens from n-1-length patterns by depth first search.
:param extensionsOfP: it is n-1 length patterns list.
:type extensionsOfP: list
:return: None
"""
for x in range(len(extensionsOfP)-1):
extensionsOfPx = set()
for y in range(x+1,len(extensionsOfP)):
tspxy = self.__tsList[extensionsOfP[x]] & self.__tsList[extensionsOfP[y]]
PTL = self.__calculatePTL(tspxy)
if len(PTL) > 0:
if type(extensionsOfP[x]) == str:
pattern = (extensionsOfP[x], extensionsOfP[y])
self._localPeriodicPatterns__finalPatterns[pattern] = PTL
self.__tsList[pattern] = tspxy
extensionsOfPx.add(pattern)
else:
px = [item for item in extensionsOfP[x]]
py = [item for item in extensionsOfP[y]]
pattern = set(px + py)
self._localPeriodicPatterns__finalPatterns[tuple(pattern)] = PTL
self.__tsList[tuple(pattern)] = tspxy
extensionsOfPx.add(tuple(pattern))
if extensionsOfPx:
self.__LPPMDepthSearch(list(extensionsOfPx))
def __convert(self, value: Union[int, float, str]) -> Union[int, float]:
"""
to convert the type of user specified minSup value
:param value: user specified minSup value
:type value: int or float or str
:return: converted type
:rtype: int or float
"""
if type(value) is int:
value = int(value)
if type(value) is float:
value = (len(self.__Database) * value)
if type(value) is str:
if '.' in value:
value = float(value)
value = (len(self.__Database) * value)
else:
value = int(value)
return value
[docs]
@deprecated("It is recommended to use 'mine()' instead of 'mine()' for mining process. Starting from January 2025, 'mine()' will be completely terminated.")
def startMine(self) -> None:
"""
Mining process start from here. This function calls createTSlist and generateLPP.
"""
self.mine()
[docs]
def mine(self) -> None:
"""
Mining process start from here. This function calls createTSlist and generateLPP.
"""
self._localPeriodicPatterns__startTime = _ab._time.time()
self._localPeriodicPatterns__finalPatterns = {}
self.__creatingItemSets()
self._localPeriodicPatterns__maxPer = self.__convert(self._localPeriodicPatterns__maxPer)
self._localPeriodicPatterns__maxSoPer = self.__convert(self._localPeriodicPatterns__maxSoPer)
self._localPeriodicPatterns__minDur = self.__convert(self._localPeriodicPatterns__minDur)
self.__createTSlist()
self.__generateLPP()
self._localPeriodicPatterns__endTime = _ab._time.time()
process = _ab._psutil.Process(_ab._os.getpid())
self._localPeriodicPatterns__memoryRSS = float()
self._localPeriodicPatterns__memoryUSS = float()
self._localPeriodicPatterns__memoryUSS = process.memory_full_info().uss
self._localPeriodicPatterns__memoryRSS = process.memory_info().rss
[docs]
def getMemoryUSS(self) -> float:
"""
Total amount of USS memory consumed by the mining process will be retrieved from this function
:return: returning USS memory consumed by the mining process
:rtype: float
"""
return self._localPeriodicPatterns__memoryUSS
[docs]
def getRuntime(self) -> float:
"""
Calculating the total amount of runtime taken by the mining process
:return: returning total amount of runtime taken by the mining process
:rtype: float
"""
return self._localPeriodicPatterns__endTime - self._localPeriodicPatterns__startTime
[docs]
def getPatternsAsDataFrame(self) -> _ab._pd.DataFrame:
"""
Storing final local periodic patterns in a dataframe
:return: returning local periodic patterns in a dataframe
:rtype: pd.DataFrame
"""
dataFrame = {}
data = []
for a, b in self._localPeriodicPatterns__finalPatterns.items():
pat = str()
for i in a:
pat = pat + i + ' '
data.append([pat, b])
dataFrame = _ab._pd.DataFrame(data, columns=['Patterns', 'PTL'])
return dataFrame
[docs]
def save(self, outFile: str) -> None:
"""
Complete set of local periodic patterns will be loaded in to an output file
:param outFile: name of the output file
:type outFile: csv file
:return: None
"""
self._localPeriodicPatterns__oFile = outFile
writer = open(self._localPeriodicPatterns__oFile, 'w+')
for x, y in self._localPeriodicPatterns__finalPatterns.items():
pat = str()
for i in x:
pat = pat + i + '\t'
pat = pat + ":"
for i in y:
pat = pat + str(i) + '\t'
patternsAndPTL = pat.strip()
writer.write("%s \n" % patternsAndPTL)
[docs]
def getPatterns(self) -> Dict[Union[Tuple[str, ...], str], Set[Tuple[int, int]]]:
"""
Function to send the set of local periodic patterns after completion of the mining process
:return: returning frequent patterns
:rtype: dict
"""
return self._localPeriodicPatterns__finalPatterns
[docs]
def printResults(self) -> None:
"""
This function is used to print the results
"""
print("Total number of Local Periodic Patterns:", len(self.getPatterns()))
print("Total Memory in USS:", self.getMemoryUSS())
print("Total Memory in RSS", self.getMemoryRSS())
print("Total ExecutionTime in ms:", self.getRuntime())
if __name__ == '__main__':
_ap = str()
if len(_ab._sys.argv) == 5 or len(_ab._sys.argv) == 6:
if len(_ab._sys.argv) == 6:
_ap = LPPMDepth(_ab._sys.argv[1], _ab._sys.argv[3], float(_ab._sys.argv[4]), _ab._sys.argv[5])
_ap.mine()
_ap.save(_ab._sys.argv[2])
if len(_ab._sys.argv) == 5:
_ap = LPPMDepth(_ab._sys.argv[1], _ab._sys.argv[2], float(_ab._sys.argv[3]), _ab._sys.argv[4])
_ap.mine()
print("Total number of Local Periodic Patterns:", len(_ap.getPatterns()))
print("Total Memory in USS:", _ap.getMemoryUSS())
print("Total Memory in RSS", _ap.getMemoryRSS())
print("Total ExecutionTime in seconds:", _ap.getRuntime())
else:
print("Error! The number of input parameters do not match the total number of parameters provided")