def CentralAndNorm(self):
"""centralize and normalize the total dataframe
"""
from centralize_outlier import centralizeOutlier
from normalize import linearNormalize
df_to_process = self.total_dt.drop(['mcc'], axis=1)
centralizeOutlier(df_to_process, devs=3)
df_to_process = linearNormalize(df_to_process, ceil=1.0, floor=(-1.0))
mcc = df(self.total_dt['mcc'])
self.total_dt = pd.concat([df_to_process, mcc], axis=1)
def NormAndCentralize(self, centralized=True):
mcc = self.all_set[:, 0]
df_to_process = self.all_set[:, 1:]
df_to_process = pd.DataFrame(df_to_process)
if centralized == True:
centralizeOutlier(df_to_process, devs=self.devs)
df_to_process = linearNormalize(df_to_process, ceil=1.0, floor=-1.0)
df_to_process = df_to_process.values
self.all_set = np.column_stack((mcc, df_to_process))
def preSvcData(self, print_it=False):
"""
collect the data for SVC,
centralize and normalize all the columns except for mcc,
user can choose to print the table by setting print_it to True.
"""
df_to_process = self.svc_sheet.drop(['mcc'], axis=1)
centralizeOutlier(df_to_process, devs=self.devs)
df_to_process = linearNormalize(df_to_process, ceil=1.0, floor=-1.0)
mcc = pd.DataFrame(self.svc_sheet['mcc'])
self.svc_sheet = pd.concat([df_to_process, mcc], axis=1)
def preAllData(self, print_it=False, centralized=True):
"""
for the whole data set:
centralize and normalize all the columns except for mcc,
user can choose to print the table by setting print_it to True.
"""
df_to_process = self.all_sheet.drop(['mcc'], axis=1)
if centralized == True:
centralizeOutlier(df_to_process, devs=self.devs)
df_to_process = linearNormalize(df_to_process, ceil=1.0, floor=-1.0)
mcc = pd.DataFrame(self.all_sheet['mcc'])
self.all_sheet = pd.concat([df_to_process, mcc], axis=1)
def processMccEner(self, centralized=True, normed=True):
f = h5py.File(self.h5_path)
dt = f[self.all_path][()]
f.close()
self.whole_dt = pd.DataFrame(dt)
self.whole_dt.columns = self.rep_ener_columns
self.dropWrongMcc()
if centralized == True:
df_to_process = self.whole_dt.drop(['mcc'], axis=1)
centralizeOutlier(df_to_process, devs=self.devs)
if normed == True:
df_to_process = linearNormalize(df_to_process, ceil=1.0, floor=-1.0)
mcc = pd.DataFrame(self.whole_dt['mcc'])
self.whole_dt = pd.concat([mcc, df_to_process], axis=1)