def initPredict(self):
datas = csv.readCSVFile(self.trainFilename, ',')
if (datas is None):
return False
try:
trainFeatures = datas[self.featureColumns]
except Exception:
print("One or multiple columns beetween: ",
", ".join(self.featureColumns), " doesn't exits")
try:
self.classes = datas[self.classifyColumn]
except Exception:
print("the classify column ", self.classifyColumn, "doesn't exits")
predictFeature = csv.readCSVFile(self.predictFilename, ',')
if (predictFeature is None):
return False
try:
predictFeature = predictFeature[self.featureColumns]
except Exception:
print("One or multiple columns beetween: ",
", ".join(self.featureColumns), " doesn't exits")
self.predictFeatures = self.normalizePredictFile(
trainFeatures, predictFeature)
self.predictFeatures.insert(
0, 'theta0', [1.0 for _ in range(self.predictFeatures.shape[0])])
if (self.predictFeatures is not None):
self.initPredictDone = True
return self.initPredictDone
def predictAll(self):
predicts = []
if (self.initPredictDone is False):
print(
"The function init must return True before call the function train"
)
return None
if (self.predictFilename is None):
print("Add the predict filename in constructor")
return None
datas = csv.readCSVFile("thetas.csv", ',')
if (datas is None):
return False
try:
thetas = datas[['theta0'] + self.featureColumns]
except Exception:
print("Error in thetas File")
return None
classNames = self.classes.unique()
for values in self.predictFeatures.values:
results = [
self.predict(lineThetas, values.tolist())
for lineThetas in thetas.values
]
predicts.append(classNames[results.index(max(results))])
dataframe = {}
dataframe.update({self.classifyColumn: predicts})
return pd.DataFrame(dataframe)
def main():
vlen = len(sys.argv)
if (vlen == 2 or (vlen == 3 and sys.argv[1] == "-all")):
datas = csv.readCSVFile(sys.argv[vlen - 1], ',')
if (datas is None):
sys.exit(1)
subjectDatas = csv.dropColumns(datas, csv.notArithmetiqueSubjects)
if (subjectDatas is None):
sys.exit(1)
subjectNames, subjectsByHouse = csv.getSubjectValueByHouse(datas)
fig = plt.figure()
elements = []
size = len(subjectNames)
sqrtSize = math.sqrt(size)
sqrtSize = int(sqrtSize) if (sqrtSize
== int(sqrtSize)) else int(sqrtSize + 1)
if (vlen == 3):
for subjectIndex in range(size):
elements.append(
fig.add_subplot(sqrtSize, sqrtSize, subjectIndex + 1))
elements[subjectIndex].hist(subjectsByHouse[subjectIndex],
label=csv.houseNames,
color=csv.colors)
elements[subjectIndex].set_title(subjectNames[subjectIndex])
elements[subjectIndex].legend()
else:
elements.append(fig.add_subplot(1, 1, 1))
elements[0].hist(subjectsByHouse[10],
label=csv.houseNames,
color=csv.colors)
elements[0].set_title(subjectNames[10])
elements[0].legend()
plt.show()
else:
print('Error script : python histogram.py [-all] file.')
def main():
if len(sys.argv) == 2:
datas = csv.readCSVFile(sys.argv[1], ',')
if (datas is None):
sys.exit(1)
subjectDatas = csv.dropColumns(datas, csv.notArithmetiqueSubjects)
if (subjectDatas is None):
sys.exit(1)
sns.relplot(x="Astronomy",
y="Defense Against the Dark Arts",
hue="Hogwarts House",
hue_order=csv.houseNames,
palette=csv.colors,
data=subjectDatas)
plt.show()
else:
print('Error script : python scatter_.py file.')
def main():
vlen = len(sys.argv)
if (vlen == 2 or (vlen == 3 and sys.argv[1] == "-all")):
datas = csv.readCSVFile(sys.argv[vlen - 1], ',')
if (datas is None):
sys.exit(1)
subjectDatas = csv.dropColumns(datas, csv.notArithmetiqueSubjects)
if (subjectDatas is None):
sys.exit(1)
# print(subjectDatas.describe())
print("Describe general:\n{}\n".format(describe(subjectDatas)))
if (vlen == 3):
for house in csv.houseNames:
print("Describe {}:\n{}\n".format(house, describe(subjectDatas.loc[subjectDatas['Hogwarts House'].isin({house})])))
else:
print('Error script : python describe.py [-all] file.')
def initTrain(self):
datas = csv.readCSVFile(self.trainFilename, ',')
if (datas is None):
return False
try:
self.features = self.normalize(datas[self.featureColumns])
self.features.insert(0, 'theta0',
[1.0 for _ in range(self.features.shape[0])])
except Exception:
print("One or multiple columns beetween: ",
", ".join(self.featureColumns), " doesn't exits")
try:
self.classes = datas[self.classifyColumn]
except Exception:
print("the classify column ", self.classifyColumn, "doesn't exits")
if (all(variable is not None
for variable in [self.features, self.classes])):
self.initTrainDone = True
return self.initTrainDone