def main_thread():
global main_array
global score_array
temp_return = user_score.main_thread(file_path)
temp_return = nom.min_max_normalize(temp_return)
main_array[0][1] = temp_return['3544']
main_array[1][1] = temp_return['48117']
main_array[2][1] = temp_return['58744']
main_array[3][1] = temp_return['60606']
main_array[4][1] = temp_return['60619']
main_array[5][1] = temp_return['60708']
main_array[6][1] = temp_return['60728']
main_array[7][1] = temp_return['60797']
main_array[8][1] = temp_return['61143']
temp_return = frac.main_thread(file_path)
temp_return = nom.min_max_normalize(temp_return)
main_array[0][2] = temp_return['3544']
main_array[1][2] = temp_return['48117']
main_array[2][2] = temp_return['58744']
main_array[3][2] = temp_return['60606']
main_array[4][2] = temp_return['60619']
main_array[5][2] = temp_return['60708']
main_array[6][2] = temp_return['60728']
main_array[7][2] = temp_return['60797']
main_array[8][2] = temp_return['61143']
temp_return = time.main_thread(file_path)
temp_return = nom.min_max_normalize(temp_return)
main_array[0][3] = temp_return['3544']
main_array[1][3] = temp_return['48117']
main_array[2][3] = temp_return['58744']
main_array[3][3] = temp_return['60606']
main_array[4][3] = temp_return['60619']
main_array[5][3] = temp_return['60708']
main_array[6][3] = temp_return['60728']
main_array[7][3] = temp_return['60797']
main_array[8][3] = temp_return['61143']
X_matrix = np.array(main_array)
Y_matrix_t = np.array(score_array)
Y_matrix = Y_matrix_t.transpose()
res_vector_t = np.dot(
np.dot(np.linalg.inv(np.dot(X_matrix.transpose(), X_matrix)),
X_matrix.transpose()), Y_matrix)
res_vector = list(res_vector_t.transpose())
for i in range(len(res_vector)):
res_vector[i] = float(res_vector[i])
print(res_vector)
print(main_array[0][0] * res_vector[1] + main_array[0][1] * res_vector[2] +
main_array[0][2] * res_vector[3] + main_array[0][3] * res_vector[4] -
res_vector[0])
return res_vector
def json_read():
global main_array
f = open(file_path_style, 'r', encoding='utf-8')
data = json.load(f)
data = nom.min_max_normalize(data)
main_array[0][4] = data['3544']
main_array[1][4] = data['48117']
main_array[2][4] = data['58744']
main_array[3][4] = data['60606']
main_array[4][4] = data['60619']
main_array[5][4] = data['60708']
main_array[6][4] = data['60728']
main_array[7][4] = data['60797']
main_array[8][4] = data['61143']
def main_thread():
temp_return = userScore.main_thread(file_path)
temp_return = normal.min_max_normalize(temp_return)
temp_value_list = list(temp_return.values())
diff_return.append(temp_value_list)
temp_return = fraction.main_thread(file_path)
temp_return = normal.min_max_normalize(temp_return)
temp_value_list = list(temp_return.values())
diff_return.append(temp_value_list)
temp_return = firstUpload.main_thread(file_path)
temp_return = normal.min_max_normalize(temp_return)
temp_value_list = list(temp_return.values())
diff_return.append(temp_value_list)
temp_return = time.main_thread(file_path)
temp_return = normal.min_max_normalize(temp_return)
temp_value_list = list(temp_return.values())
diff_return.append(temp_value_list)
temp_return = compare.main_thread(file_path)
temp_return = normal.min_max_normalize(temp_return)
temp_value_list = list(temp_return.values())
diff_return.append(temp_value_list)
f = open('SampleCodingStyleOutput.json', 'r',
encoding='utf-8') # codingStyle这个速度，还是直接读文件比较好
data = json.load(f)
temp_return = normal.min_max_normalize(data)
temp_value_list = list(temp_return.values())
diff_return.append(temp_value_list)
a = np.asarray(diff_return)
# print(a)
# correlation_matrix = np.corrcoef(a, rowvar=True)
# print(correlation_matrix.round(2))
positive_relations = []
negative_relations = []
corr_matrix_list = []
p_matrix_list = []
for i in range(len(a)):
tmp_0 = []
tmp_1 = []
for j in range(len(a)):
corr, p = pearsonr(a[i], a[j])
tmp_0.append(corr)
tmp_1.append(p)
if p < 0.01 and i != j:
if corr > 0.9 and [j, i] not in positive_relations:
positive_relations.append([i, j])
# list_merge(positive_relations)
if corr < -0.9 and [j, i] not in negative_relations:
negative_relations.append([i, j])
# list_merge(negative_relations)
corr_matrix_list.append(tmp_0)
p_matrix_list.append(tmp_1)
corr_matrix = np.asarray(corr_matrix_list)
p_matrix = np.asarray(p_matrix_list)
print(corr_matrix.round(2))
print(p_matrix.round(2))
positive_negative = [positive_relations, negative_relations]
print(positive_negative)
return positive_negative
def main_thread():
temp_return = firstgit.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][0] = temp_return[case_id]
temp_return = final.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][1] = temp_return[case_id]
temp_return = avg.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][2] = temp_return[case_id]
temp_return = frac.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][3] = temp_return[case_id]
temp_return = slope.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][4] = temp_return[case_id]
temp_return = time.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][5] = temp_return[case_id]
temp_return = com.main_thread(file_path)
temp_return = nomal.min_max_normalize(temp_return)
temp_case_list = temp_return.keys()
for case_id in temp_case_list:
diff_return[int(case_id)][6] = temp_return[case_id]
for i in range(3000):
num = aver(diff_return[i])
for j in range(7):
diff_return[i][j] -= num
matrix_diff_origin = numpy.array(diff_return)
# use numpy to calculate covariance(协方差) matrix of the matrix and view each row as a var
cov_matrix = numpy.cov(matrix_diff_origin, rowvar=False)
# calculate matrix eigenvalues of the matrix
# eig_matrix_list[0]特征值，[1]特征向量
eig_matrix_list = numpy.linalg.eig(cov_matrix)
eig_matrix_value = list(numpy.linalg.eigvals(cov_matrix))
main_var_index = eig_matrix_value.index(max(
eig_matrix_value)) # eig_matrix_list[0].index(max(eig_matrix_value))
# have Eigenvector of the matrix
eig_vector_matrix = numpy.array(list(eig_matrix_list[1][main_var_index]))
# transpose the matrix
trans_eig_vec_matrix = eig_vector_matrix.transpose(
) # numpy.transpose(eig_vector_matrix)
# 两矩阵相乘
res_matrix = list(numpy.dot(matrix_diff_origin, trans_eig_vec_matrix))
for i in range(3000):
res_matrix[i] = float(res_matrix[i])
min_val = min(res_matrix)
max_val = max(res_matrix)
sub_val = max_val - min_val
for i in range(len(res_matrix)):
res_matrix[i] = round((res_matrix[i] - min_val) / sub_val, 2)
# print(res_matrix)
return res_matrix