def refreshBars():
setMainTable(Tables.Tables(getMainWindow()).Table)
getMainWindow().SpecialTools = SpecialTools.SpecialTools(getMainWindow())
if uni.tableType in ["2", "6", "8", "9"]:
from Bars import PlayerBar
getMainWindow().PlayerBar = PlayerBar.PlayerBar(getMainWindow())
getMainWindow().addToolBar(Mt.TopToolBarArea,
getMainWindow().PlayerBar)
getMainWindow().Menu.refreshForTableType()
def predict_length(self,vector,course_list):
tb = Tables()
tb.read_data("grade2.csv", "student.csv")
vector=self.courses_to_numeric(vector,True)
lst = np.empty(tb.studyTable.shape[1])
lst[:] = np.nan
for i in range(len(course_list)):
if course_list[i] in tb.courseList:
lst[tb.courseList.index(course_list[i])] = vector[i]
lst = lst.reshape(1,-1)
lst = tb.study_imp.transform(lst)
result = self.study_length.predict(lst)
return np.round(result[0])
def predict_dropout(self,vector,course_list):
tb = Tables()
tb.read_data("grade2.csv", "student.csv")
vector=self.courses_to_numeric(vector,True)
lst = np.empty(tb.dropoutTable.shape[1])
lst[:] = np.nan
for i in range(len(course_list)):
if course_list[i] in tb.courseList:
lst[tb.courseList.index(course_list[i])] = vector[i]
lst = lst.reshape(1,-1)
lst = tb.dropout_imp.transform(lst)
result = self.dropout.predict(lst)
if result[0] == 1: return True
else: return False
def predict_gpa(self,vector,course_list):
tb = Tables()
tb.read_data("grade2.csv", "student.csv")
vector=self.courses_to_numeric(vector,True)
lst = np.empty(tb.graduationTable.shape[1])
lst[:] = np.nan
for i in range(len(course_list)):
if course_list[i] in tb.courseList:
lst[tb.courseList.index(course_list[i])] = vector[i]
lst = lst.reshape(1,-1)
lst = tb.graduation_imp.transform(lst)
result = self.gpa.predict(lst)
if result<0: result=result*-1
return "%.2f"%result
def Output():
db.create_all()
if request.method == 'POST':
userdata = Tables(request.form["Username"], request.form["psw"])
users = Tables.query.filter_by(
Username=request.form["Username"]).first()
if users is not None:
message = "User name is already existing.Register with new user name"
return render_template("Registration.html", message1=message)
db.session.add(userdata)
db.session.commit()
message = "Registartion is successful"
return render_template("Registration.html", message=message)
# result = request.form
# print(result['Username'])
else:
return render_template("Registration.html")
def create_new_model(self, grade_file, student_file, predict_function, algorithm_name, parameters, course_name, semester):
tb = Tables()
tb.read_data(grade_file, student_file)
if predict_function=='gpa':
if algorithm_name=='linear':
return ML.linear_regression(tb.graduationTable, tb.graduationLabel, parameters)
elif predict_function=='dropout':
if algorithm_name=='logistic':
return ML.logistic_regression(tb.dropoutTable, tb.dropoutLabel, parameters)
elif algorithm_name=='svm':
return ML.svm(tb.dropoutTable, tb.dropoutLabel, parameters)
elif algorithm_name=='mlp':
return ML.mlp(tb.dropoutTable, tb.dropoutLabel, parameters)
elif predict_function=='course_grade':
##get required table without target label
c = [0,0.5,1,1.5,2,2.5,3,3.5,4]
x = tb.courseTable.copy()
class_index = tb.courseList.index(course_name.lower())
y = x[:,class_index] ## class label
x = np.delete(x,class_index,1) ## remove class column from input data
## change y label with class value (eg. if y[0] = 3.5 then it become 7th class)
for i in range(y.size):
index = c.index(y[i])
y[i] = index
if algorithm_name=='logistic':
return ML.logistic_regression(x,y,parameters)
elif algorithm_name=='svm':
return ML.svm(x,y,parameters)
elif algorithm_name=='mlp':
return ML.mlp(x,y,parameters)
elif predict_function=='study_length':
if algorithm_name=='linear':
return ML.linear_regression(tb.studyTable, tb.studyLabel, parameters)
elif algorithm_name=='mlp_regressor':
return ML.mlp_regressor(tb.studyTable, tb.studyLabel, parameters)
def predict_course_grade(self,vector,course_list,pcourse_name):
tb = Tables()
tb.read_data("grade2.csv", "student.csv")
if (pcourse_name not in tb.courseList):
return (pcourse_name+" is not available anymore")
model=None
q_course = "SELECT path \
FROM models \
WHERE gradeFile='%s' AND studentFile='%s' AND function='%s' AND course='%s' AND isDefault='%d'" % ("grade2.csv","student.csv","course_grade",pcourse_name,1)
db = MySQLdb.connect("localhost",mysql_user,password,"gpa_db_4" )
cursor = db.cursor()
try:
cursor.execute(q_course)
path = cursor.fetchone()
model = joblib.load(path[0])
except IOError as e:
print("Model file doesn't exist.")
db.close()
if model==None:
return ("There is no available model file for" + pcourse_name)
vector=self.courses_to_numeric(vector,True) ##convert course grades to numeric ones(eg. AA=4, CB = 2.5)
lst = np.empty(tb.courseTable.shape[1])
lst[:] = np.nan
for i in range(len(course_list)):
if course_list[i] in tb.courseList:
lst[tb.courseList.index(course_list[i])] = vector[i]
lst = lst.reshape(1,-1)
lst = tb.course_imp.transform(lst)
class_index = tb.courseList.index(pcourse_name)
lst = np.delete(lst,class_index,1) ## remove class column from input data
result = model.predict(lst)
result = self.courses_to_numeric(result,False)
return result[0]
def __init__(self, value):
self.value = value
self.tables= Tables(value)
self.left = None
self.right = None
self.height = 0 #altura
