import matplotlib.pyplot as plt
os.environ["TF_CPP_MIN_LOG_LEVEL"]='2'
## PRI
from ML_funcs import unpickle,create_model
## setting enviroment and data
MODEL_DIR = "F:/data_eeg/result_model_LA"
dic_file = 'F:/data_eeg/all_pkl/result_LA_pkl.pkl'
BATCH_SIZE = 128
EPOCHS = 20
num_Classes = 4
checkpoint = ModelCheckpoint(filepath=os.path.join(MODEL_DIR,'M_{epoch:03d}_l_{loss:.3f}_vl_{val_loss:.3f}.hdf5'),save_best_only=True)
all_data = unpickle(dic_file)
train_data, test_data, train_labels_one_hot, test_labels_one_hot = train_test_split(all_data['data'], all_data['labels'])
#print('before',train_labels_one_hot.shape, test_labels_one_hot.shape)
train_data = np.array(train_data)
test_data = np.array(test_data)
print(train_data.shape,test_data.shape)
train_data = np.reshape(train_data, (train_data.shape[0],600,1))
test_data = np.reshape(test_data, (test_data.shape[0],600,1)) # do it if need to change data struct
print(train_data.shape,test_data.shape)
# train_labels_one_hot = keras.utils.to_categorical(train_labels_one_hot, num_Classes)
# test_labels_one_hot = keras.utils.to_categorical(test_labels_one_hot, num_Classes)
train_data = train_data.astype('float32')
test_data = test_data.astype('float32')
print(train_data.shape, len(train_labels_one_hot))
from ML_funcs import unpickle, minmax
from sklearn.decomposition import PCA
import sklearn.preprocessing
import matplotlib.pyplot as plt
import random
import keras
import pickle as pkl
dic_file_fa = 'F:/Insula-Gcamp6/record/record_split_by_behav/50%_nogo_hit_trials.pkl'
dic_file_cr = 'F:/Insula-Gcamp6/record/record_split_by_behav/50%_nogo_miss_trials.pkl'
result_path = 'C:/Users/manggny/Desktop/record_all/'
num_Classes = 2
#checkpoint = ModelCheckpoint(filepath=os.path.join(MODEL_DIR,'M_{epoch:03d}_l_{loss:.3f}_vl_{val_loss:.3f}.hdf5'),save_best_only=True)
data_fa = unpickle(dic_file_fa)
data_cr = unpickle(dic_file_cr)
ticks = 28
start = 0
time_acc = np.zeros(ticks) # 1 for 0.2s, all 3s (-2s~1s)
loop_time = 1000
time_step = 25
lower = min(np.alen(data_fa), np.alen(data_cr))
# for i in range(np.alen(data_fa)):
# data_fa[i,:] = sklearn.preprocessing.scale(data_fa[i,:],axis=0) #,feature_range=(-1,1)
# for i in range(np.alen(data_cr)):
# data_cr[i, :] = sklearn.preprocessing.scale(data_cr[i, :], axis=0) #,feature_range=(-1,1)
for loop in range(loop_time):
idx_data_fa = list(range(np.alen(data_fa)))
import numpy as np
import pickle as pkl
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from sklearn.preprocessing import StandardScaler
from sklearn.cluster import MiniBatchKMeans
from matplotlib import style
style.use('ggplot')
from ML_funcs import unpickle
dic_name = 'C:/Users/Administrator/Desktop/EEGDATA/result_sub_features_4labels.pkl'
data = unpickle(dic_name)
print(np.array(data['data']).shape)
result_dic = {'data':[],'labels':data['labels']}
data = np.array(data['data'])
tr,ti,elec = np.shape(data)
result = np.zeros((tr,7,elec))
for i in range(elec):
data_dummy = data[:,:,i]
scal = StandardScaler()
scal.fit(data_dummy)
s_data = scal.transform(data_dummy)
p = PCA(n_components = 7)
from ML_funcs import unpickle, mean_range, create_model
import numpy as np
import pickle as pkl
import sklearn.preprocessing
all_data_path = 'C:/Users/Administrator/Desktop/EEGDATA/result_boot_r_0702.pkl'
data_all = unpickle(all_data_path)
print(data_all['labels'].count(0), data_all['labels'].count(1),
data_all['labels'].count(2), data_all['labels'].count(3))
trial_num = len(data_all['data'])
datas = []
labels = []
print(data_all['data'][1].shape)
for i in range(trial_num):
_, col = data_all['data'][i].shape
print(data_all['data'][i].shape)
if (data_all['labels'][i] == 1): # or (data_all['labels'][i] == 2):
print(data_all['labels'][i])
for k in range(col):
data_all['data'][i][:, k] = sklearn.preprocessing.scale(
data_all['data'][i][:, k])
datas.append(data_all['data'][i])
labels.append(1)
elif (data_all['labels'][i] == 0):
print(data_all['labels'][i])
for k in range(col):
data_all['data'][i][:, k] = sklearn.preprocessing.scale(
data_all['data'][i][:, k])
datas.append(data_all['data'][i])
from typing import List
from ML_funcs import unpickle
import matplotlib.pyplot as plt
import numpy as np
from numpy.core._multiarray_umath import ndarray
result_path = 'C:/Users/Administrator/Desktop/EEGDATA/result_svm/'
path = 'C:/Users/Administrator/Desktop/EEGDATA/all_txts'
dic_path = 'C:/Users/Administrator/Desktop/EEGDATA/all_pkl/svm_result_acc_r_boot_0702.pkl'
data = unpickle(dic_path)
print(data)
mean_acc = []
for i in range(35):
mean_acc.append(np.mean(data[:,i]))
print(np.mean(data[:,i]))
dummy = mean_acc[:]
sort_idx =[]
print(mean_acc)
for i in range(35):
best_acc = max(dummy)
for k in range(35):
if dummy[k] == best_acc and best_acc > 0:
sort_idx.append(k)
dummy[k] = -1
break
print(sort_idx)
#print(mean_acc)