def generateDataset():
ResultHeadings, ResultData = getDataFromResults("results.csv")
PitStopsHeadings, PitStopsData, PitStopsDict = getPitStops("pitStops.csv")
LapTimesHeadings, LapTimesData, LapTimesDict, RaceLapDict = getLapTimes(
"lapTimes.csv")
InsertFirstLapChange(ResultHeadings, ResultData, RaceLapDict)
keys = PitStopsDict.keys()
InsertPitStopsTime(ResultHeadings, ResultData, PitStopsDict)
ResultData.insert(0, ResultHeadings)
saveLListAsCSV("PreprocessedDataset1", ResultData)
def generateDataset():
ResultHeadings, ResultData = getDataFromResults("results.csv")
PitStopsHeadings, PitStopsData, PitStopsDict = getPitStops("pitStops.csv")
LapTimesHeadings, LapTimesData, LapTimesDict, RaceLapDict = getLapTimes(
"lapTimes.csv")
InsertFirstLapChange(ResultHeadings, ResultData, RaceLapDict)
keys = PitStopsDict.keys()
InsertPitStopsTime(ResultHeadings, ResultData, PitStopsDict)
_data_needed = []
for data in ResultData:
if ((data[7] >= "A") & (data[7] <= "Z")):
continue
else:
_data_needed.append(data)
_data_needed.insert(0, ResultHeadings)
saveLListAsCSV("PreprocessedDataset1", _data_needed)
def getCircuitSpecificStatisticsOfFirstLapChange(headings):
"""
for circuit specific statistics
"""
headings.append("DriverId")
headings.append("DriverName")
headings.append("CircuitId")
headings.append("CircuitName")
headings.append("FirstLapChangeList")
headings.append("FirstLapChangeMin")
headings.append("FirstLapChangeMax")
headings.append("FirstLapChangeFirstQuartile")
headings.append("FirstLapChangeMedian")
headings.append("FirstLapChangeThirdQuartile")
headings.append("FirstLapChangeMean")
headings.append("FirstLapChangeSd")
_ret_list = []
circuitRaceList = getCircuitRaceList("races.csv")
PreprocessedDataHeadings, PreprocessedDataset = getPreprocessedData(
"PreprocessedDataset1.csv")
circuitNameDict = getRecentTracksName()
driverIDNameDict = getDriverIDName("drivers.csv")
# temperary data structure
# {driverId: {circuitId: [first lap change]}}
_dict_dict_list = defaultdict(lambda: defaultdict(lambda: []))
for data in PreprocessedDataset:
circuitKeys = circuitRaceList.keys()
for circuitKey in circuitKeys:
if (data[0] in circuitRaceList[circuitKey]):
_dict_dict_list[data[1]][circuitKey].append(data[9])
# print _dict_dict_list["4"]["32"]
driversKeys = sorted(_dict_dict_list.keys(), key=lambda _key: int(_key))
for driversKey in driversKeys:
_driverID = driversKey
_driverName = driverIDNameDict[_driverID]
_circuitID = ""
_circuitName = ""
_FirstLapChangeList = ""
_FirstLapChangeMin = 0.0
_FirstLapChangeMax = 0.0
_FirstLapChangeFirstQuartile = 0.0
_FirstLapChangeMedian = 0.0
_FirstLapChangeThirdQuartile = 0.0
_FirstLapChangeMean = 0.0
_FirstLapChangeSd = 0.0
driverCircuitsKeys = sorted(_dict_dict_list[driversKey].keys(),
key=lambda _key: int(_key))
for driverCircuitsKey in driverCircuitsKeys:
# _tempSumOfFirstLapChange = 0.0;
_tempContainer = []
_circuitID = driverCircuitsKey
_circuitName = circuitNameDict[_circuitID]
for firstLapChangeVal in _dict_dict_list[driversKey][
driverCircuitsKey]:
if (firstLapChangeVal != "-999"):
_FirstLapChangeList += (firstLapChangeVal + ",")
_change = int(firstLapChangeVal)
_tempContainer.append(_change)
_tempContainer.sort(key=lambda _key: int(_key))
if (len(_tempContainer) != 0):
_FirstLapChangeMin = _tempContainer[0]
_FirstLapChangeMax = _tempContainer[-1]
_FirstLapChangeFirstQuartile = _tempContainer[
len(_tempContainer) / 4]
_FirstLapChangeMedian = _tempContainer[len(_tempContainer) / 2]
_FirstLapChangeThirdQuartile = _tempContainer[
len(_tempContainer) / 4 * 3]
_FirstLapChangeMean = np.sum(
np.array(_tempContainer)) / len(_tempContainer)
_FirstLapChangeSd = np.sqrt(
np.sum((
(np.array(_tempContainer) - _FirstLapChangeMean)**2)) /
len(_tempContainer))
_ret_list.append([
_driverID, _driverName, _circuitID, _circuitName,
_FirstLapChangeList, _FirstLapChangeMin, _FirstLapChangeMax,
_FirstLapChangeFirstQuartile, _FirstLapChangeMedian,
_FirstLapChangeThirdQuartile, _FirstLapChangeMean,
_FirstLapChangeSd
])
# print _FirstLapChangeList
# if(len(_tempContainer) != 0):
# _FirstLapChangeMin=_tempContainer[0]
# _FirstLapChangeMax=_tempContainer[-1]
# _FirstLapChangeFirstQuartile = _tempContainer[len(_tempContainer)/4]
# _FirstLapChangeMedian=_tempContainer[len(_tempContainer)/2]
# _FirstLapChangeThirdQuartile = _tempContainer[len(_tempContainer)/4 *3]
# _FirstLapChangeMean=np.sum(np.array(_tempContainer))/len(_tempContainer)
# _FirstLapChangeSd=np.sqrt(np.sum(((np.array(_tempContainer)-_FirstLapChangeMean)**2))/len(_tempContainer))
# _ret_list.append([_driverID, _FirstLapChangeList, _FirstLapChangeMin, _FirstLapChangeMax, _FirstLapChangeFirstQuartile,
# _FirstLapChangeMedian, _FirstLapChangeThirdQuartile, _FirstLapChangeMean, _FirstLapChangeSd])
_ret_list.insert(0, headings)
saveLListAsCSV("PreprocessedDataset3", _ret_list)
def createDataFrame(headings, data):
headings = ["driverId", "ranking", "circuits"]
data.insert(0,headings)
saveLListAsCSV("DriversTopPerformingCircuits", data)
_listOfList.append([ConstructorIDName[data[2]],\
data[1], \
DriverName[data[1]],\
circuitIDNameDict[_circuitID],\
_totalpoints[data[1]]/len(_driverYear[data[1]]),\
returnCategoryNumber(rank)
])
return {"listofList": _listOfList}
if __name__ == '__main__':
dataset = getRaceDriverConstRankPts("results.csv")
DPS = getDriverIDPointsConstructor(dataset)
# print DPS["listofList"]
# newArray = np.array(DPS["listofList"])[]
DCAPR = getDriverIDConstructorAveragePointsRanking(dataset)
# print DCAPR
FilteredDataset = filterDataAccordingToRecentDrivers(DPS["listofList"])
FilteredDCAPR = filterDataAccordingToRecentDrivers2(DCAPR["listofList"])
saveLListAsCSV("PreprocessedDataset4", FilteredDataset)
saveLListAsCSV("PreprocessedDataset5", FilteredDCAPR)
headings = []
createDataFrame(headings, getDriverTopPerformingCircuits())
_ret_list_of_list = list(listOfList)
_ret_list_of_list.insert(0, headings)
return _ret_list_of_list
elif (typeOfDataFrame == "similarityList"):
headings = ["circuitId1", "circuitId2", "n"]
for i in xrange(1000):
headings.append("P" + str(i + 1))
_ret_list_of_list = listOfList
_ret_list_of_list.insert(0, headings)
return _ret_list_of_list
if __name__ == "__main__":
print "start"
"""
obtain the frequency of active circuit and plot a scatter plot
"""
circuitDict_InfoList = getCircuitInfoList("new_track_data.csv")
# print len(circuitDict_InfoList["20"])
# shiftRightNTimes(4, circuitDict_InfoList["20"])
# optimisedSimilarity(circuitDict_InfoList["20"], circuitDict_InfoList["18"])
covarianceMatrixHeadings, covarianceMatrix, similarityList = getCovarianceMatrix(
circuitDict_InfoList)
CovarianceMatrix = createDataFrame(covarianceMatrix, "covarianceMatrix",
covarianceMatrixHeadings)
SimilarityInfoList = createDataFrame(similarityList, "similarityList", [])
saveLListAsCSV("CovarianceMatrixOfCircuits", CovarianceMatrix)
saveLListAsCSV("SimilarityInfoList", SimilarityInfoList)
print "end"