script_r415_process_record.py

#process all record
import pickle
import signal_processing as sig_proc

dir_name = '../data/r415/'

img_ext = '.png'
save_img = False
show = False
save_obj = False

#signal filtering parameter
low_cut = 3e2
high_cut = 3e3

#spike finding parameters
b_spike = 6
a_spike = 20
spike_thresh = -4

threshold_template = 4 #distance from which it's acceptable to put spike in class
base_name = 'r415_'
record_name = ['130926', '131008', '131009', '131011', '131016', '131017', '131018', '131021', '131023', '131025',
               '131030', '131101', '131118', '131129']
record_data = {}
with open(dir_name + 'templates', 'rb') as my_file:
    all_chan_templates = pickle.load(my_file)

sp = sig_proc.Signal_processing(save_img, show, img_ext)

#for record in record_name:	
record = record_name[0]
print('----- processing record: ' + record + ' -----')
signal = sp.load_m(dir_name + base_name + record + '.mat', 'd')#load multichannel signal
fs = float(sp.load_m(dir_name + 'fech.mat', 'sampFreq')) #load sample frequency

fsignal = sp.signal_mc_filtering(signal, low_cut, high_cut, fs)

all_chan_spikes_values = []
all_chan_spikes_times = []
all_chan_spikes_classes = []
all_chan_clusters = []
#create cluster for storing classed spikes inside
print('### create cluster storing spikes_values and spikes_time ###')
for chan in range(len(all_chan_templates)):
    all_chan_clusters.append([])
    for gpe in all_chan_templates[chan]:
        all_chan_clusters[chan].append(sig_proc.Spikes_cluster(gpe.template, gpe.number + 1))
for chan in range(fsignal.shape[0]):
    print('\n\n--- processing chan: ' + str(chan + 1) + ' ---')
    s = fsignal[chan]
    spikes_values, spikes_time = sp.find_spikes(s, a_spike, b_spike, spike_thresh)
    spikes_values = sp.smooth_spikes(spikes_values, 3)
    print('spikes found: ' + str(spikes_values.shape[0]))
    spikes_classes = sp.classify_spikes(spikes_values, spikes_time, all_chan_clusters[chan], 2 * threshold_template)

    all_chan_spikes_classes.append(spikes_classes)
    all_chan_spikes_times.append(spikes_time)
    all_chan_spikes_values.append(spikes_values)
    sp.plot_signal_spikes_classified(spikes_time, spikes_classes, base_name + '_' + record + '_' + str(chan) + '_')

# unclass_spikes=[]
# for i in range(len(spikes_classes)):
# if spikes_classes[i]==-1:
# unclass_spikes.append(spikes_values[i])
# sp.plot_spikes(np.array(unclass_spikes),20,'chan_'+str(chan))

#plt.show()
record_data[record] = {'spikes_values': all_chan_spikes_values, 'spikes_time': all_chan_spikes_times,
                       'spikes_classes': all_chan_spikes_classes, 'clusters': all_chan_clusters}

if save_obj:
    with open(dir_name + 'data_processed', 'wb') as my_file:
        my_pickler = pickle.Pickler(my_file)
        my_pickler.dump(record_data)

print('\n\n#################')
print('####   END   ####')