Page MenuHomec4science
Files F83476840

compute_srf.py
No OneTemporary
Actions

File Metadata

Created
Tue, Sep 17, 08:44

compute_srf.py
View Options

import numpy as np
import os
from scipy.interpolate import CubicSpline, interp1d
from tqdm import tqdm
import linecache
import tracemalloc
DATA_DIR = "../data/extracted_data/"
import pickle
from multiprocessing import Pool
import sys
# Signal freqeuncy variables
FREQ = 128
MULTIPLIER_FREQ = 50 #Final freq: 128*20 = 2.56 KHz
def open_file(file_name):
file_name_full = os.path.join(DATA_DIR, os.path.basename(file_name))
data = {}
with open(file_name_full,"rb") as f:
data = pickle.load(f)
f.flush()
f.close()
return data
def find_num_events_ADC(signal, nBits, frequency_multiplier = MULTIPLIER_FREQ ,hist = 5, original_bits = 11):
#ADC stats
up_samp_signal = upsample_uniform_signal(signal, frequency_multiplier)
delta = 2**original_bits
dV = (delta)/(2**nBits)
hist = (hist/100)*dV
min_val = -delta//2
#init value, first sample (we assume we always sample the nearest level at start) and ADC status
v_0 = up_samp_signal['v'][0]
lowTh = min_val+((v_0-min_val)//dV)*dV
highTh = lowTh + dV
num_events_ADC = 0
for val in tqdm(up_samp_signal['v'],desc = f"ADC with {nBits} bits"):
#print(f"Value: {val}, time: {time}, low_th = {lowTh - hist}, high_th = { highTh + hist}")
if val > highTh + hist or val < lowTh - hist:
lowTh = min_val+((val-min_val)//dV)*dV #Delta from the bottom: (val-min_val)//dV*dV then compute the actual level summin min_val
highTh = lowTh + dV
num_events_ADC+=1
return num_events_ADC
def ADC(signal, nBits, frequency_multiplier = MULTIPLIER_FREQ ,hist = 5, original_bits = 11):
#ADC stats
up_samp_signal = upsample_uniform_signal(signal, frequency_multiplier)
delta = 2**original_bits
dV = (delta)/(2**nBits)
hist = (hist/100)*dV
min_val = -delta//2
events = {'t':[],'v':[]}
#init value, first sample (we assume we always sample the nearest level at start) and ADC status
v_0 = up_samp_signal['v'][0]
lowTh = min_val+((v_0-min_val)//dV)*dV
highTh = lowTh + dV
events['t'].append(up_samp_signal['t'][0]/frequency_multiplier)
events['v'].append(int(lowTh if v_0-lowTh < highTh - v_0 else highTh))
i = 0
for val in tqdm(up_samp_signal['v'],desc = f"ADC with {nBits} bits"):
#print(f"Value: {val}, time: {time}, low_th = {lowTh - hist}, high_th = { highTh + hist}")
if val > highTh + hist or val < lowTh - hist:
direction = 1 if val > highTh else -1
lowTh = min_val+((val-min_val)//dV)*dV #Delta from the bottom: (val-min_val)//dV*dV then compute the actual level summin min_val
highTh = lowTh + dV
events['t'].append(up_samp_signal['t'][i]/frequency_multiplier)
events['v'].append(int(lowTh if direction == 1 else highTh))
i+=1
if events['t'][-1] < up_samp_signal['t'][-1]/frequency_multiplier:
#final value, last sample (we assume we always sample the nearest level at end)
v_end = up_samp_signal['v'][-1]
lowTh = min_val+((v_end-min_val)//dV)*dV
highTh = lowTh + dV
events['t'].append(up_samp_signal['t'][-1]/frequency_multiplier)
events['v'].append(int(lowTh if v_end-lowTh < highTh - v_end else highTh))
return events
def upsample_uniform_signal(signal,multiplier):
out_signal = {}
upsampled = []
v = signal['v']
last_new_t = (len(v)-1)*multiplier
t = np.arange(0,last_new_t+1,multiplier,dtype = np.int64)
new_base = np.arange(0,last_new_t,1,dtype = np.int64)
f = interp1d(t, v)
upsampled = f(new_base)
out_signal = {'t':new_base,'v':upsampled}
return out_signal
def prof_one_lvl(b):
tracemalloc.start()
original_num_points = 0
EB_num_points = 0
snapshot_start = tracemalloc.take_snapshot()
for f_name in os.listdir(DATA_DIR):
print(f"File {f_name}")
data = open_file(f_name)
EB_num_points += find_num_events_ADC(data, b)
original_num_points += len(data['t'])
data = None
snapshot_end = tracemalloc.take_snapshot()
top_stats = snapshot_end.statistics('lineno')
top_stats_diff = snapshot_end.compare_to(snapshot_start, 'lineno')
print("\n\n########################################################\n\n")
print("[ Top 10 memory usage]")
for stat in top_stats[:10]:
print(stat)
print("\n\n########################################################\n\n")
print("[ Top 10 differences ]")
for stat in top_stats_diff[:10]:
print(stat)
print("\n\n########################################################\n\n")
print(f'{b} Bits, SRF = {(original_num_points-EB_num_points)/original_num_points}\n\n')
snapshot_start = snapshot_end
SRF = (original_num_points-EB_num_points)/original_num_points
print(f'{b} Bits, SRF = {SRF}')
with open('../results/SRFs_3.txt','a+') as f:
f.write(f'{b} Bits, SRF = {SRF}\n')
def process_one_level(lvl):
original_num_points = 0
EB_num_points = 0
for f_name in os.listdir(DATA_DIR):
tracemalloc.start()
tracemalloc.stop()
print(f"File {f_name}")
data = open_file(f_name)
EB_num_points += find_num_events_ADC(data, lvl)
original_num_points += len(data['t'])
data = None
SRF = (original_num_points-EB_num_points)/original_num_points
print(f'{lvl} Bits, SRF = {SRF}')
with open('../results/SRFs_3.txt','a+') as f:
f.write(f'{lvl} Bits, SRF = {SRF}\n')
def main():
cores = int(sys.argv[2])
with Pool(cores) as p:
if sys.argv[1] == "proces":
p.map(process_one_level, list(range(2,11)))
elif sys.argv[1] == "profile":
p.map(prof_one_lvl, list(range(2,11)))
else:
print(f"YOU STUPID FUCK!")
exit
if __name__ == "__main__":
main()

Event Timeline

c4science · Help