import cv2.ml as ml
from MachineLearn.Classes.data import Data
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.experiment import Experiment
SAMPLES_PER_CLASS = 150
PATH_TO_SAVE_FEATURES = '../../GLCM_FILES/EXP_07/'
NUMBER_OF_ROUNDS = 50
MIN_DECIMATION = 50
MAX_DECIMATION = 100
EXPERIMENT = 7
NBITS = 8
NATT = 24
oExp = Experiment()
# basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
# basemask = np.array([12, 20, 22])
basemask = np.array(range(1, 25))
svmVectors = []
basemask = basemask - 1
for M in range(MIN_DECIMATION, MAX_DECIMATION + 1):
oDataSet = DataSet()
base = np.loadtxt(PATH_TO_SAVE_FEATURES + "FEATURES_M{}_CM8b_TH199.txt".format(M), usecols=basemask, delimiter=",")
classes = np.loadtxt(PATH_TO_SAVE_FEATURES + "FEATURES_M{}_CM8b_TH199.txt".format(M), dtype=object, usecols=24,
delimiter=",")
for x, y in enumerate(base):
oDataSet.add_sample_of_attribute(np.array(list(np.float32(y)) + [classes[x]]))
oDataSet.attributes = oDataSet.attributes.astype(float)
from MachineLearn.Classes.experiment import Experiment
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.data import Data
import numpy as np
import cv2
oExp = Experiment()
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23, 25])
svmVectors = []
qtd_per_class = np.array([82, 93, 140, 60])
###################################################################################################################################
#basemask = basemask-1
#for i in range(1):
#oDataSet = DataSet()
#base = np.loadtxt("GLCM_FILES/M1_CM8b_RE_ELIAS.txt", usecols=(x for x in range(24)), delimiter=",")
#classes = np.loadtxt("GLCM_FILES/M1_CM8b_RE_ELIAS.txt",dtype=object, usecols=(24), delimiter=",")
#classes = np.matrix(classes)
#base = np.array(np.hstack((base,classes.T)))
#base[base == "c1_p1"] = "c1"
#base[base == "c2_p1"] = "c2"
#base[base == "c3_p1"] = "c3"
#base[base == "c3_p2"] = "c3"
#base[base == "c3_p3"] = "c3"
#base[base == "c3_p4"] = "c3"
#base[base == "c4_p1"] = "c4"
#base[base == "c4_p2"] = "c4"
#base[base == "c4_p3"] = "c4"
#base[base == "c4_p4"] = "c4"
#unique, counts = np.unique(base[:,-1], return_counts=True)
ATT_NUMBER = 24
DECIMATION = 17
DECIMATION_MIN = 1
DECIMATION_MAX = 1
NUMBER_OF_ROUNDS = 50
CM_BIT_MIN = 8
CM_BIT_MAX = 8
CM_BIT = 8
TH_MIN = 199
TH_MAX = 199
R = 17
ROUND = 0
SHOW = True
oExp = Experiment.load(
"OBJECTS/EXP_{:02d}/ACC_M{}-{}_{}_CM{}-{}b_TH{}-{}_ATT{}.gzip".format(
EXPERIMENT_NUMBER, DECIMATION_MIN, DECIMATION_MAX, NUMBER_OF_ROUNDS,
CM_BIT_MIN, CM_BIT_MAX, TH_MIN, TH_MAX, ATT_NUMBER))
# oExp2 = Experiment.load(
# "OBJECTS/EXP_{:02d}/ACC_M{}-{}_{}_CM{}-{}b_TH{}-{}_ATT{}.gzip".format(EXPERIMENT_NUMBER, 50,
# 100,
# NUMBER_OF_ROUNDS, CM_BIT_MIN, CM_BIT_MAX,
# TH_MIN, TH_MAX,
# ATT_NUMBER))
# oExp = Experiment()
# oExp.experimentResults = oExp1.experimentResults[:-1] + oExp2.experimentResults
# oExp.experimentDescription = oExp1.experimentDescription[:-1] + oExp2.experimentDescription
# oExp.length = oExp1.length + oExp2.length -1
print(oExp.show_in_table())
if SHOW:
for j, i in enumerate(oExp.experimentResults):
import os
from MachineLearn.Classes.experiment import Experiment
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.data import Data
import numpy as np
import cv2
M = 17
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
TH = 199
supportVectors = np.zeros((7, 50))
oExp = Experiment.load("../OBJECTS/EXP_06/ACC_M17-17_50_CM2-7b_TH199-199_ATT10.gzip")
for k in range(6):
oDataSet = oExp.experimentResults[k]
for i, j in enumerate(oDataSet.dataSet):
j.save_model("tmp.txt")
svm = cv2.SVM()
svm.load("tmp.txt")
supportVectors[k, i] = svm.get_support_vector_count()
oExp = Experiment.load("../OBJECTS/EXP_06/ACC_M1-100_50_CM8-8b_TH199-199_ATT10.gzip")
oDataSet = oExp.experimentResults[16]
for i, j in enumerate(oDataSet.dataSet):
j.save_model("tmp.txt")
svm = cv2.SVM()
svm.load("tmp.txt")
supportVectors[6, i] = svm.get_support_vector_count()
os.remove("tmp.txt")
import cv2
import numpy as np
from MachineLearn.Classes.data import Data
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.experiment import Experiment
SAMPLES_PER_CLASS = 50
PATH_TO_SAVE_FEATURES = 'GLCM_FILES/EXP_04/'
NUMBER_OF_ROUNDS = 50
MIN_THRESHOLD = 190
MAX_THRESHOLD = 200
oExp = Experiment()
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
svmVectors = []
basemask = basemask - 1
#
#
# for TH in range(MIN_THRESHOLD, MAX_THRESHOLD + 1):
# oDataSet = DataSet()
# base = np.loadtxt(PATH_TO_SAVE_FEATURES + "FEATURES_M1_CM8b_TH{}.txt".format(TH), usecols=basemask, delimiter=",")
# classes = np.loadtxt(PATH_TO_SAVE_FEATURES + "FEATURES_M1_CM8b_TH{}.txt".format(TH), dtype=object, usecols=24, delimiter=",")
# for x, y in enumerate(base):
# oDataSet.addSampleOfAtt(np.array(list(np.float32(y)) + [classes[x]]))
# oDataSet.atributes = oDataSet.atributes.astype(float)
# oDataSet.normalizeDataSet()
# for j in range(NUMBER_OF_ROUNDS):
# print j
# oData = Data(4, 13, samples=50)
import cv2
import numpy as np
from MachineLearn.Classes.data import Data
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.experiment import Experiment
SAMPLES_PER_CLASS = 50
PATH_TO_SAVE_FEATURES = '../../GLCM_FILES/EXP_04/'
NUMBER_OF_ROUNDS = 50
MIN_BITS = 2
MAX_BITS = 8
DECIMATION = 14
oExp = Experiment()
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
svmVectors = []
basemask = basemask - 1
for n_bits in range(MIN_BITS, MAX_BITS + 1):
oDataSet = DataSet()
base = np.loadtxt(
PATH_TO_SAVE_FEATURES +
"FEATURES_M{}_CM{}b_TH198.txt".format(DECIMATION, n_bits),
usecols=basemask,
delimiter=",")
classes = np.loadtxt(
PATH_TO_SAVE_FEATURES +
"FEATURES_M{}_CM{}b_TH198.txt".format(DECIMATION, n_bits),
dtype=object,
import cv2
import numpy as np
from MachineLearn.Classes.data import Data
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.experiment import Experiment
SAMPLES_PER_CLASS = 50
PATH_TO_SAVE_FEATURES = 'GLCM_FILES/EXP_04/'
NUMBER_OF_ROUNDS = 50
MIN_DECIMATION = 1
MAX_DECIMATION = 100
oExp = Experiment()
basemask = np.array(range(1, 25))
svmVectors = []
basemask = basemask - 1
for M in range(MIN_DECIMATION, MAX_DECIMATION + 1):
oDataSet = DataSet()
base = np.loadtxt(PATH_TO_SAVE_FEATURES +
"FEATURES_M{}_CM8b_TH198.txt".format(M),
usecols=basemask,
delimiter=",")
classes = np.loadtxt(PATH_TO_SAVE_FEATURES +
"FEATURES_M{}_CM8b_TH198.txt".format(M),
dtype=object,
usecols=24,
delimiter=",")
for x, y in enumerate(base):
from MachineLearn.Classes.experiment import Experiment
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.data import Data
import numpy as np
import cv2
oExp = Experiment()
M = 1
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
accuracy = 0
nVectors = 10000
PATH_TO_SAVE_FEATURES = '../../GLCM_FILES/EXP_04/'
TH = 198
ROUNDS = 10000
basemask = basemask - 1
best = 0
for i in range(1):
oDataSet = DataSet()
base = np.loadtxt(PATH_TO_SAVE_FEATURES +
"FEATURES_M{}_CM8b_TH{}.txt".format(M, TH),
usecols=basemask,
delimiter=",")
classes = np.loadtxt(PATH_TO_SAVE_FEATURES +
"FEATURES_M{}_CM8b_TH{}.txt".format(M, TH),
dtype=object,
usecols=24,
delimiter=",")
for x, y in enumerate(base):
oDataSet.add_sample_of_attribute(
np.array(list(np.float32(y)) + [classes[x]]))
import numpy as np
from MachineLearn.Classes.data import Data
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.experiment import Experiment
SAMPLES_PER_CLASS = 150
PATH_TO_SAVE_FEATURES = '../../GLCM_FILES/EXP_06/'
NUMBER_OF_ROUNDS = 50
MIN_THRESHOLD = 181
MAX_THRESHOLD = 194
TH_STEP = 1
EXPERIMENT = 6
oExp = Experiment()
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
svmVectors = []
basemask = basemask - 1
for TH in range(MIN_THRESHOLD, MAX_THRESHOLD + 1, TH_STEP):
oDataSet = DataSet()
base = np.loadtxt(PATH_TO_SAVE_FEATURES +
"FEATURES_M1_CM8b_TH{}.txt".format(TH),
usecols=basemask,
delimiter=",")
classes = np.loadtxt(PATH_TO_SAVE_FEATURES +
"FEATURES_M1_CM8b_TH{}.txt".format(TH),
dtype=object,
usecols=24,
delimiter=",")
import numpy as np
from MachineLearn.Classes.experiment import Experiment
EXPERIMENT_NUMBER = 6
ATT_NUMBER = 10
MIN_DECIMATION = 1
MAX_DECIMATION = 100
NUMBER_OF_ROUNDS = 50
CM_BITS = 8
TH = 199
oExp = Experiment.load(
"../OBJECTS/EXP_{:02d}/ACC_M{}-{}_{}_CM{}-{}b_TH{}-{}_ATT{}.gzip".format(EXPERIMENT_NUMBER, MIN_DECIMATION,
MAX_DECIMATION,
NUMBER_OF_ROUNDS, CM_BITS, CM_BITS, TH, TH,
ATT_NUMBER))
results = np.zeros((MAX_DECIMATION - MIN_DECIMATION + 1, 3))
for i, oDataSet in enumerate(oExp.experimentResults):
metrics = oDataSet.get_general_metrics()
results[i, 0] = metrics[0][0, -1] - (metrics[1][0, -1]*.75)
results[i, 1] = metrics[0][0, -1]
results[i, 2] = metrics[0][0, -1] + metrics[1][0, -1]
results = results * 100
possibles_a = []
possibles_b = []
possibles_c = []
avg_minus = np.min(results[:3, 0])
from MachineLearn.Classes.experiment import Experiment
from MachineLearn.Classes.data_set import DataSet
from MachineLearn.Classes.data import Data
import numpy as np
import cv2
M = 1
basemask = np.array([1, 2, 5, 9, 15, 16, 17, 21, 22, 23])
TH = 198
ROUNDS = 10000
oExp = Experiment.load(
"../OBJECTS/EXP_04/BEST_MODEL_{}_CROSSWALK_M{}_TH{}.txt".format(
ROUNDS, M, TH))
oDataSet = oExp.experimentResults[0]
bestIndexies = []
bestP1 = 0
bestP2 = 0
nvector = 1000000
i = 0
results = np.zeros((4, 4, 2))
for i, j in enumerate(oDataSet.dataSet):
j.save_model("tmp.txt")
svm = cv2.SVM()
svm.load("tmp.txt")
results1 = svm.predict_all(
np.float32(oDataSet.attributes[j.Testing_indexes]))
results2 = svm.predict_all(
np.float32(oDataSet.attributes[j.Training_indexes]))
acc = j.get_metrics()[0][-1]