def Transform(self, dc: DataContainer, store_folder=None, store_key=None):
if dc.IsEmpty():
return dc
new_dc = deepcopy(dc)
array = new_dc.array
array -= self._interception
array /= self._slop
array = np.nan_to_num(array)
new_dc.array = array
new_dc.UpdateFrame()
if store_folder is not None and store_key is not None:
assert(len(store_key) > 0)
self.SaveNormalDataContainer(new_dc, store_folder, store_key)
return new_dc
def __init__(self):
super(AalenAdditive, self).__init__(AalenAdditiveFitter(),
self.__class__.__name__)
#
# class Weibull(BaseFitter):
# def __init__(self):
# super(Weibull, self).__init__(WeibullAFTFitter(), self.__class__.__name__)
if __name__ == '__main__':
import numpy as np
model = CoxPH()
print(model.name)
# model = AalenAdditive()
# print(model.name)
train_dc = DataContainer()
train_dc.Load(r'..\..\Demo\train.csv',
event_name='status',
duration_name='time')
model.Fit(train_dc)
result = model.Summary()
print(result)
# model.Save(r'..\..\Demo')
#
# model_new = AalenAdditive()
# model_new.Load(r'..\..\Demo')
# model_new.Summary()
assignments = np.array(
(dc.df.shape[0] // self.k + 1) * list(range(1, self.k + 1)))
assignments = assignments[:dc.df.shape[0]]
for i in range(1, self.k + 1):
ix = assignments == i
train_dc = deepcopy(dc)
train_dc.df = df.loc[~ix]
train_dc.UpdateData()
val_dc = deepcopy(dc)
val_dc.df = df.loc[ix]
val_dc.UpdateData()
yield train_dc, val_dc
if __name__ == '__main__':
from SA.Fitter import MyCox
dc = DataContainer()
dc.Load(r'C:\Users\yangs\Desktop\Radiomics_pvp_hcc_os_top20_train.csv',
event_name='status',
duration_name='time')
print(dc)
fitter = MyCox()
cv = CrossValidation()
cv.Generate(dc)
def Transform(self, dc: DataContainer, store_folder=None, store_key=None):
assert(dc.GetFeatureName() == self.sub_features)
if store_folder is not None and store_key is not None:
dc.Save(os.path.join(store_folder, '{}_reduced_features.csv'.format(store_key)))
return dc
def SaveNormalDataContainer(self, dc: DataContainer, store_folder, store_key):
dc.Save(os.path.join(store_folder, '{}_normalized_{}_feature.csv'.format(self._name, store_key)))
"and divided by it. Then the feature vector was mapped to an unit vector. "
NormalizerMinMax = Normalizer('MinMax', unit_description, MinMaxNormFunc)
def ZNormalizeFunc(array):
return np.std(array, axis=0), np.mean(array, axis=0)
z_description = "We applied the normalization on the feature matrix. For each feature vector, we calculated the mean " \
"value and the standard deviation. Each feature vector was subtracted by the mean value and was divided " \
"by the standard deviation. After normalization process, each vector has zero center and unit standard " \
"deviation. "
NormalizerZscore = Normalizer('Zscore', z_description, ZNormalizeFunc)
def MeanNormFunc(array):
return np.max(array, axis=0) - np.min(array, axis=0), np.mean(array, axis=0)
z_0_description = "We applied the normalization on the feature matrix. Each feature vector was subtracted by the mean " \
"value of the vector and was divided by the length of it. "
NormalizerMean = Normalizer('Mean', z_0_description, MeanNormFunc)
if __name__ == '__main__':
file_path = r'C:\Users\yangs\Desktop\Radiomics_pvp_hcc_os_top20_train.csv'
dc = DataContainer()
dc.Load(file_path, event_name='status', duration_name='time')
normal = NormalizerNone
new_dc = normal.Fit(dc)
print(new_dc)
new_dc = normal.Transform(dc)
print(new_dc)