from time import time, strftime
from datetime import date
from sklearn.metrics import roc_curve, roc_auc_score, precision_recall_curve, average_precision_score
from components.grading.grading_pipelines import pipeline_prediction
from components.grading.roc_curve import roc_curve_single, roc_curve_multi, calc_curve_bootstrap, display_bootstraps, plot_vois
from components.utilities.load_write import load_excel
if __name__ == '__main__':
# Arguments
choice = 'Isokerays'
datapath = r'/media/dios/dios2/3DHistoData'
# datapath = r'X:/3DHistoData'
arguments = arg.return_args(datapath,
choice,
pars=arg.set_surf_loo,
grade_list=arg.grades_cut)
arguments.save_path = arguments.save_path
combinator = np.mean
arguments.binary_model = 'LOG'
# LOGO for 2mm samples
if choice == '2mm':
arguments.split = 'logo'
arguments.train_regression = True
groups, _ = load_excel(arguments.grade_path, titles=['groups'])
groups = groups.flatten()
elif choice == 'Isokerays' or choice == 'Isokerays_sub':
arguments.train_regression = False
arguments.split = 'logo'
metric,
groups=pat_groups)
print('Results for grades: ' + args.grades_used)
print("Parameters are:\n", pars)
for i in range(len(pars)):
print(pars[i])
if __name__ == '__main__':
start_time = time()
# Arguments
dataset_name = '2mm'
data_path = r'/media/dios/dios2/3DHistoData'
arguments = arg.return_args(data_path,
dataset_name,
grade_list=arg.grades_cut)
# !Decline PCA usage!
#arguments.use_PCA = False
arguments.standardization = 'standardize'
#arguments.alpha = 1.0
arguments.split = 'logo'
arguments.n_pars = 5
loss_function = mean_squared_error
# Groups
if dataset_name == '2mm':
groups, _ = load_excel(arguments.grade_path, titles=['groups'])
groups = groups.flatten()
import components.utilities.listbox as listbox
from components.processing.voi_extraction_pipelines import pipeline_subvolume_mean_std
from components.utilities.load_write import find_image_paths
from components.grading.roc_curve import roc_curve_multi, roc_curve_single
from scripts.run_lbp_features_vois import pipeline_lbp
from scripts.run_pca_regression import pipeline_prediction
if __name__ == '__main__':
# Arguments
choice = 'Isokerays'
data_path = r'Y:\3DHistoData'
arguments_p = arg_process.return_args(data_path, choice)
arguments_g = arg_grading.return_args(data_path,
choice,
pars=arg_grading.set_90p,
grade_list=arg_grading.grades)
# Path to image stacks
arguments_p.data_path = r'U:\PTA1272\Isokerays_PTA'
# LOGO for 2mm samples
if choice == '2mm':
arguments_g.split = 'logo'
groups = np.array([
1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 8, 8, 9, 9, 10, 10, 11, 11,
12, 12, 13, 13, 14, 14, 15, 16, 16, 17, 18, 19, 19
]) # 2mm, 34 patients
else:
groups = None
# Use listbox to select samples (Result is saved in listbox.file_list)
listbox.GetFileSelection(arguments_p.data_path)
"""Calculates MRELBP features from mean+std images for given parameters (see return_args(pars))."""
import components.grading.args_grading as arg
from components.utilities import listbox
from components.grading.grading_pipelines import pipeline_lbp
if __name__ == '__main__':
# Arguments
choice = '2mm'
datapath = r'X:\3DHistoData'
arguments = arg.return_args(datapath, choice, pars=arg.set_90p_2m_cut, grade_list=arg.grades_cut)
arguments.save_path = r'X:\3DHistoData\Grading\LBP\\' + choice
# Use listbox (Result is saved in listbox.file_list)
listbox.GetFileSelection(arguments.image_path)
# Call pipeline
for k in range(len(arguments.grades_used)):
pars = arguments.pars[k]
grade_selection = arguments.grades_used[k]
print('Processing with parameters: {0}'.format(grade_selection))
pipeline_lbp(arguments, listbox.file_list, pars, grade_selection)
from components.grading.grading_pipelines import pipeline_lbp, pipeline_prediction
from components.grading.roc_curve import roc_curve_single, roc_curve_multi, calc_curve_bootstrap, plot_vois
from components.utilities.load_write import load_excel
if __name__ == '__main__':
# Arguments
start_time = time()
warnings.filterwarnings("ignore", category=DeprecationWarning)
dataset_name = '2mm'
data_path = r'/media/dios/dios2/3DHistoData'
combinator = np.mean
# Get arguments as namespace
arguments = arg.return_args(data_path,
dataset_name,
pars=arg.set_surf_loo,
grade_list=arg.grades_cut)
#arguments = arg.return_args(data_path, dataset_name, pars=arg.set_FS, grade_list=arg.grades_cut)
if dataset_name == '2mm':
arguments.image_path = '/media/santeri/data/MeanStd_2mm_augmented'
arguments.train_regression = True
arguments.split = 'logo'
groups, _ = load_excel(arguments.grade_path, titles=['groups'])
groups = groups.flatten()
elif dataset_name == 'Isokerays' or dataset_name == 'Isokerays_sub':
arguments.image_path = '/media/santeri/data/MeanStd_4mm_augmented'
arguments.train_regression = True
#arguments.n_subvolumes = 9
groups, _ = load_excel(arguments.grade_path, titles=['groups'])