PACK_in-vivo/script1_line-length_analysis.py

# -*- coding: utf-8 -*-
"""
Created on Wed Oct  9 17:29:15 2019

@author: kleisp
"""

import os
os.chdir('C:\EAfiles\EAdetection')

import core.ea_management as eam
from core.helpers import open_hdf5
from glob import glob
#import phases_ictalInter as pii
#import core.ea_analysis as eana

import pandas as pd
import numpy as np


os.chdir('C:\EAfiles\EAdetection')
reload(eam)


#%% function for getting line length directly from resampled raw data

def get_linelength(filename,ll_dict_controls, recID): 
  excelpath = 'C:\EAfiles\Excel\\startstop_PK_HCc.xlsx' #insert the directory for an excel file containing the start and stop times for each recording (first column: recID, second column: start time, third column stop time)
  df = pd.read_excel(excelpath) #reading an excel file with start and stop times (s) of a recording
  df.set_index('recID', inplace=True) #recID
  ll_dict_controls[recID]['start'] = int(df.start[recID]) #start time
  ll_dict_controls[recID]['stop'] = int(df.stop[recID]) #stop time
  sr=500 #sampling rate of downsampled data
  data_dict=open_hdf5(filename,group=None,read_maskedarr=False)
  data=data_dict['data']['trace']
  reccut=data[int(ll_dict_controls[recID]['start']*sr):int(ll_dict_controls[recID]['stop']*sr)] #cut data according to start and stop times
  N_datapoints=len(reccut) 
  duration=N_datapoints/sr
  linelength = np.sum(np.abs(np.diff(reccut))) #line length is the sum of subsequent datapoints -->derivative
  linelength_s=linelength/duration
  ll_dict_controls[recID]['linelength/second']=linelength_s
  ll_dict_controls[recID]['linelength']=linelength
  
#%% get linelength directly from h5 (downsampled data)

os.chdir(r'C:\EAfiles\DATA\h5files') #enter directory, where all the needed h5 files are
files = glob(r'C:\EAfiles\DATA\h5files\*.h5') #takes all h5 files from this folder
hfile = files
lines = np.arange(0, len(hfile)) #as many lines as there are h5 files
ll_dict_controls={}
ll_dict_controls_d={}

for i in lines:
    recID = os.path.basename(hfile[i]).split('__')[0] #getting the IDs
    print (i)
    print (recID)
    
    filename=os.path.basename(hfile[i])
    ll_dict_controls[recID]= {}
    get_linelength(filename,ll_dict_controls, recID)

#%% creating arrays with data 
recIDs = np.array(ll_dict_controls.keys())

flavour = 'linelength/second'
ii_linelengths = np.array([ll_dict_controls[recID][flavour] for recID in ll_dict_controls.keys()])

#%% create excel file

os.chdir('C:\EAfiles\RESULTS')

df_sr=pd.DataFrame(columns=
    ['recID', 
    'linelength/s'])

df_sr['recID'] = recIDs
df_sr['linelength/s'] = ii_linelengths
df_sr.to_excel('test_linelengths.xlsx') #name the file