import pickle
# General python libraries
import scipy.signal as sig
import numpy as np
# Plotting Libraries
import matplotlib.pyplot as plt
import seaborn as sns
#Do some cleanup of the plotting space
plt.close('all')
sns.set_context('paper')
sns.set_style('white')
sns.set(font_scale=4)
# Misc libraries
import copy
import itertools
import scipy.stats as stats
import ipdb
#%% Initial
# Now we set up our DBSpace environment
ClinFrame = ClinVect.CFrame(norm_scales=True)
#BRFrame = BRDF.BR_Data_Tree(preFrame='Chronic_Frame_2019.pickle')
BRFrame = pickle.load(
open('/home/virati/Dropbox/Data/Chronic_Frame_2019.pickle', "rb"))
readout = DSV.DMD_RO(BRFrame, ClinFrame)
readout.default_run()
import itertools
import pickle
from DBSpace.readout import DSV
#%%
#This sets up our clinical frame for the regression
print('Loading Clinical and BrainRadio Frames...')
ClinFrame = ClinVect.CFrame(norm_scales=True)
BRFrame = pickle.load(open('/home/virati/Chronic_Frame.pickle',"rb"))
print('Starting the readout analysis...')
analysis = DSV.ORegress(BRFrame,ClinFrame)
#%%
analysis.split_validation_set(do_split = True)
analysis.O_feat_extract()
all_pts = ['901','903','905','906','907','908']
#%%
regr_type = 'RIDGE'
test_scale = 'HDRS17'
do_detrend='Block'
ranson = True
if regr_type == 'OLSnite':
import numpy as np
#%%
ClinFrame = ClinVect.CFrame(norm_scales=True)
#ClinFrame.plot_scale(pts='all',scale='HDRS17')
#ClinFrame.plot_scale(pts=['901'],scale='MADRS')
#%%
#BRFrame = BR_Data_Tree()
#BRFrame.full_sequence(data_path='/home/virati/Chronic_Frame_july.npy')
#BRFrame.check_empty_phases()
#BRFrame = pickle.load(open('/home/virati/Chronic_Frame.pickle',"rb"))
BRFrame = pickle.load(open('/home/virati/Dropbox/Data/Chronic_FrameMay2020.pickle',"rb"))
#%%
analysis = DSV.DSV(BRFrame,ClinFrame,lim_freq=30,use_scale='HDRS17')
ENet_params = {'Alpha':(5,6),'Lambda':(0.9)}
#%%
analysis.run_EN(alpha_list=
np.linspace(40,60,100))
#%%
analysis.plot_EN_coeffs()
#%%
#aanalysis.plot_dsgn_matrix()
analysis.plot_tests()
#%%
analysis.plot_performance(ranson=True)
#%%
## MAJOR PARAMETERS for our partial biometric analysis
do_pts = ['901', '903', '905', '906', '907', '908'
] # Which patients do we want to include in this entire analysis?
#do_pts = ['901']
test_scale = 'HDRS17' # Which scale are we using as the measurement of the depression state?
''' DETRENDING
Which detrending scheme are we doing
This is important. Block goes into each patient and does zero-mean and linear detrend across time
None does not do this
All does a linear detrend across all concatenated observations. This is dumb and should not be done. Will eliminate this since it makes no sense
'''
do_detrend = 'Block'
rmethod = 'ENR_Osc'
#%%
# Now we set up our DBSpace environment
ClinFrame = ClinVect.CFrame(norm_scales=True)
BRFrame = BRDF.BR_Data_Tree()
# Run our main sequence for populating the BRFrame
BRFrame.full_sequence(
data_path=
'/home/virati/Dropbox/projects/Research/MDD-DBS/Data/Chronic_Frame_july.npy'
)
BRFrame.check_empty_phases(
) # Check to see if there are any empty phases. This should be folded into full_sequence soon TODO
readout = DSV.on_demand(BRFrame, ClinFrame, validation=0.3)
# Plotting Libraries
import matplotlib.pyplot as plt
import seaborn as sns
#Do some cleanup of the plotting space
plt.close('all')
sns.set_context('paper')
sns.set_style('white')
sns.set(font_scale=4)
# Misc libraries
import copy
import itertools
import scipy.stats as stats
#%%
ClinFrame = ClinVect.CFrame(norm_scales=True)
BRFrame = BRDF.BR_Data_Tree()
BRFrame.full_sequence(data_path='/home/virati/Chronic_Frame_july.npy')
BRFrame.check_empty_phases()
local_ro = DSV.ORegress(BRFrame, ClinFrame, trials=1, rmethod='RIDGE')
#%%
#Need to have a method that inserts models in
#need to have a method that zeros out coefficients, like right delta
#%%
# Now we want to just VIEW the coefficients, ensemble stuff, etc. Everything else from Partial_Biometric needs to be folded into ORegress class