rf_test.rar_信号_测试_脑电_脑电信号 癫痫_脑电信号特征

import os import sys import pywt from pywt import wavedec from __init__ import ap_entropy, samp_entropy import numpy as np import matplotlib from matplotlib import pyplot as plt from mnist_new import mlp, create_training_set from sklearn.ensemble import RandomForestClassifier A=[] B=[] C=[] D=[] E=[] for fl in os.listdir("../../../ALL/A/"): inp = [] path = "../../../ALL/A/" + fl txt = open(path,'r') for line in txt: feature = line.split()[0] inp.append(feature) a = np.array(inp) A.append(a) for fl in os.listdir("../../../ALL/B/"): inp = [] path = "../../../ALL/B/" + fl txt = open(path,'r') for line in txt: feature = line.split()[0] inp.append(feature) a = np.array(inp) B.append(a) for fl in os.listdir("../../../ALL/C/"): inp = [] path = "../../../ALL/C/" + fl txt = open(path,'r') for line in txt: feature = line.split()[0] inp.append(feature) a = np.array(inp) C.append(a) for fl in os.listdir("../../../ALL/D/"): inp = [] path = "../../../ALL/D/" + fl txt = open(path,'r') for line in txt: feature = line.split()[0] inp.append(feature) a = np.array(inp) D.append(a) for fl in os.listdir("../../../ALL/E/"): inp = [] path = "../../../ALL/E/" + fl txt = open(path,'r') for line in txt: feature = line.split()[0] inp.append(feature) a = np.array(inp) E.append(a) A_ = [] B_ = [] C_ = [] D_ = [] E_ = [] for x in A: coeffs = wavedec(x,'db4',level=8) A_.append(coeffs) for x in B: coeffs = wavedec(x,'db4',level=8) B_.append(coeffs) for x in C: coeffs = wavedec(x,'db4',level=8) C_.append(coeffs) for x in D: coeffs = wavedec(x,'db4',level=8) D_.append(coeffs) for x in E: coeffs = wavedec(x,'db4',level=8) E_.append(coeffs) a=[] b=[] c=[] d=[] e=[] y_a = [] y_b = [] y_c = [] y_d = [] y_e = [] f=[] y_f=[] inputs = [] outputs=[] minm = [1000000000000 for i in range(1,28)] maxm = [0 for i in range(1,28)] inp = [] out = [] for x in A_: features = [] j=0 for y in x: coef = np.array(y) energy = np.sum(coef**2) minm[j] = min(minm[j],energy) maxm[j] = max(maxm[j],energy) j=j+1 #approx_en = ap_entropy(coef,2,0.5) #minm[j] = min(minm[j],approx_en) #maxm[j] = max(maxm[j],approx_en) #j=j+1 #samp_en = samp_entropy(coef,2,0.5) #minm[j] = min(minm[j],samp_en) #maxm[j] = max(maxm[j],samp_en) #j=j+1 mean = np.mean(coef) minm[j]= min(minm[j],mean) maxm[j] = max(maxm[j],mean) j=j+1 std = np.std(coef) minm[j] = min(minm[j],std) maxm[j] = max(maxm[j],std) j=j+1 features.append(energy) #features.append(approx_en) #features.append(samp_en) features.append(mean) features.append(std) a.append(features) y_a.append(0) inputs.append(features) outputs.append("Class A") print("A done") for x in B_: features = [] j=0 for y in x: coef = np.array(y) energy = np.sum(coef**2) minm[j] = min(minm[j],energy) maxm[j] = max(maxm[j],energy) j=j+1 #approx_en = ap_entropy(coef,2,0.5) #minm[j] = min(minm[j],approx_en) #maxm[j] = max(maxm[j],approx_en) #j=j+1 #samp_en = samp_entropy(coef,2,0.5) #minm[j] = min(minm[j],samp_en) #maxm[j] = max(maxm[j],samp_en) #j=j+1 mean = np.mean(coef) minm[j]= min(minm[j],mean) maxm[j] = max(maxm[j],mean) j=j+1 std = np.std(coef) minm[j] = min(minm[j],std) maxm[j]= max(maxm[j],std) j=j+1 features.append(energy) #features.append(approx_en) #features.append(samp_en) features.append(mean) features.append(std) b.append(features) y_b.append(0) inputs.append(features) outputs.append("Class A") print("B done") for x in C_: features = [] j=0 for y in x: coef = np.array(y) energy = np.sum(coef**2) minm[j] = min(minm[j],energy) maxm[j] = max(maxm[j],energy) j=j+1 #approx_en = ap_entropy(coef,2,0.5) #minm[j] = min(minm[j],approx_en) #maxm[j] = max(maxm[j],approx_en) #j=j+1 #samp_en = samp_entropy(coef,2,0.5) #minm[j] = min(minm[j],samp_en) #maxm[j] = max(maxm[j],samp_en) #j=j+1 mean = np.mean(coef) minm[j]= min(minm[j],mean) maxm[j] = max(maxm[j],mean) j=j+1 std = np.std(coef) minm[j] = min(minm[j],std) maxm[j]= max(maxm[j],std) j=j+1 features.append(energy) #features.append(approx_en) #features.append(samp_en) features.append(mean) features.append(std) c.append(features) y_c.append(1) inputs.append(features) outputs.append("Class B") print("C done") for x in D_: features = [] j=0 for y in x: coef = np.array(y) energy = np.sum(coef**2) minm[j] = min(minm[j],energy) maxm[j] = max(maxm[j],energy) j=j+1 #approx_en = ap_entropy(coef,2,0.5) #minm[j] = min(minm[j],approx_en) #maxm[j] = max(maxm[j],approx_en) #j=j+1 #samp_en = samp_entropy(coef,2,0.5) #minm[j] = min(minm[j],samp_en) #maxm[j] = max(maxm[j],samp_en) #j=j+1 mean = np.mean(coef) minm[j]= min(minm[j],mean) maxm[j] = max(maxm[j],mean) j=j+1 std = np.std(coef) minm[j] = min(minm[j],std) maxm[j]= max(maxm[j],std) j=j+1 features.append(energy) #features.append(approx_en) #features.append(samp_en) features.append(mean) features.append(std) d.append(features) y_d.append(1) inputs.append(features) outputs.append("Class B") print("D Done") for x in E_: features = [] j=0 for y in x: coef = np.array(y) energy = np.sum(coef**2) minm[j] = min(minm[j],energy) maxm[j] = max(maxm[j],energy) j=j+1 #approx_en = ap_entropy(coef,2,0.5) #minm[j] = min(minm[j],approx_en) #maxm[j] = max(maxm[j],approx_en) #j=j+1 #samp_en = samp_entropy(coef,2,0.5) #minm[j] = min(minm[j],samp_en) #maxm[j] = max(maxm[j],samp_en) #j=j+1 mean = np.mean(coef) minm[j]= min(minm[j],mean) maxm[j] = max(maxm[j],mean) j=j+1 std = np.std(coef) minm[j] = min(minm[j],std) maxm[j]= max(maxm[j],std) j=j+1 features.append(energy) #features.append(approx_en) #features.append(samp_en) features.append(mean) features.append(std) e.append(features) f.append(features) y_e.append(2) y_f.a