from py4j.java_gateway import JavaGateway
from py4jfml.Py4Jfml import Py4jfml
import os
gateway = JavaGateway()
#percorso del file xml da valutare
str_xml = os.getcwd() + "/" + "IrisMamdani2.xml"
#Sistema fuzzy
fs = Py4jfml.load(str_xml)
#limite minimo massimo e medio SepalLength
limitiSL = [4.3, 7.9, 6.1]
#limite minimo massimo e medio SepalWidth
limitiSW = [2., 4.4, 3.2]
#limite minimo massimo e medio PetalLength
limitiPL = [1., 6.9, 3.9]
#limite minimo massimo e medio PetalWidth
limitiPW = [0.1, 2.5, 1.2]
for i in range(3):
fuzzyVarSL = fs.getVariable("SepalLength")
#imposto i valori
fuzzyVarSL.setValue(float(limitiSL[i]))
for j in range(3):
fuzzyVarSW = fs.getVariable("SepalWidth")
from py4j.java_gateway import JavaGateway
from py4jfml.Py4Jfml import Py4jfml
import matplotlib.pyplot as plt
from sklearn import datasets
from sklearn import svm
gateway = JavaGateway()
#%% load data
# import some data to play with
iris = datasets.load_iris()
#%% load FIS from FML file
fis = Py4jfml.load("IrisMamdani2.xml")
print(fis)
fis_output = []
for sample in iris.data:
fis.getVariable("SepalLength").setValue(sample[0])
fis.getVariable("SepalWidth").setValue(sample[1])
fis.getVariable("PetalLength").setValue(sample[2])
fis.getVariable("PetalWidth").setValue(sample[3])
fis.evaluate()
fis_output.append(fis.getVariable("irisClass").getValue())
#%% Train a svm from scikit-learn
clf = svm.SVC(gamma=0.001, C=100.)
clf.fit(iris.data, iris.target)
svm_output = clf.predict(iris.data)
from py4j.java_gateway import JavaGateway
from py4jfml.Py4Jfml import Py4jfml
import os
gateway = JavaGateway()
#percorso del file xml da valutare
str_xml = os.getcwd() + "/" + "IrisMamdani3.xml"
#creazione sistema fuzzy
fs1 = Py4jfml.load(str_xml)
fs2 = Py4jfml.load(str_xml)
fs3 = Py4jfml.load(str_xml)
#imposto i valori di input
#caso limite minimo
pw1 = fs1.getVariable("PetalWidth")
pw1.setValue(0.1)
#caso medio
pw2 = fs2.getVariable("PetalWidth")
pw2.setValue(2.5)
#caso limite massimo
pw3 = fs3.getVariable("PetalWidth")
pw3.setValue(1.2)
#valutazione fuzzy testlib 1
fs1.evaluate()
#valutazione fuzzy testlib 2
fs2.evaluate()
names=features + target,
dtype={target[0]: str})
fold_dict["trainingset"] = trainingset
testset = pd.read_csv("./" + fold_path + "/" + testset_filename,
header=None,
names=features + target,
dtype={target[0]: str})
fold_dict["testset"] = testset
# load FIS
fis_array = []
actual_array = []
for fis_idx in range(len(term_num)):
fis_filename = ("BEER3.txt_CV" + str(fold_idx) + "_RP" +
str(term_num[fis_idx]) + "_SP_FDTP_S.kb.xml.jfml.xml")
fis_array += [Py4jfml.load("./" + fold_path + "/" + fis_filename)]
fold_dict["fis_array"] = fis_array
# test FIS
actuals = []
fis = fis_array[fis_idx]
for i in range(len(testset)):
for f in features:
fvar = fis.getVariable(f)
if fvar is not None:
fvar.setValue(testset[f][i])
fis.evaluate()
actuals.append(str(fis.getVariable(target[0]).getValue()))
actual_array += [actuals]
fold_dict["fis_results"] = actual_array
