2022-05-11 17:32:28 +02:00
|
|
|
import random
|
|
|
|
import pandas as pd
|
2022-05-18 01:10:31 +02:00
|
|
|
import pydotplus
|
2022-05-11 17:32:28 +02:00
|
|
|
from sklearn.tree import DecisionTreeClassifier
|
|
|
|
from krata import *
|
2022-05-18 01:10:31 +02:00
|
|
|
from sklearn import metrics, tree
|
2022-05-11 17:32:28 +02:00
|
|
|
|
|
|
|
def drzewo_decyzyjne():
|
|
|
|
columns = ['plec', 'wiek', 'czas_w_pom', 'temp_w_pom', 'poziom_kurzu', 'poziom_oswietlenia', 'niebezp_towary', 'decyzja']
|
|
|
|
df = pd.read_csv("dataset.csv", header=0, sep=";", names=columns)
|
2022-05-18 01:10:31 +02:00
|
|
|
kolumny_x=['plec', 'wiek', 'czas_w_pom', 'temp_w_pom', 'poziom_kurzu', 'poziom_oswietlenia', 'niebezp_towary']
|
|
|
|
x = df[kolumny_x]
|
2022-05-11 17:32:28 +02:00
|
|
|
y = df.decyzja
|
|
|
|
#df.info()
|
|
|
|
#X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)
|
|
|
|
clf = DecisionTreeClassifier()
|
|
|
|
clf = clf.fit(x, y)
|
|
|
|
# print("Dokładność: ", metrics.accuracy_score(y_test, y_pred))
|
2022-05-18 01:10:31 +02:00
|
|
|
|
|
|
|
dot_data = tree.export_graphviz(clf, out_file=None, feature_names=kolumny_x, class_names=['0', '1'])
|
|
|
|
graph = pydotplus.graph_from_dot_data(dot_data)
|
|
|
|
graph.write_png('drzewo.png')
|
|
|
|
|
2022-05-11 17:32:28 +02:00
|
|
|
return clf
|
|
|
|
|
|
|
|
def decyzja_osoba(osoba: PoleKraty, clf: DecisionTreeClassifier):
|
|
|
|
z=[]
|
|
|
|
z.extend(random.choices([1,2], weights=[1,2], k=1)) #1 kobieta, 2 mężczyzna
|
2022-05-18 01:10:31 +02:00
|
|
|
z.append(random.randint(18, 75)) #od 55 osoba starsza
|
|
|
|
z.append(random.randint(1, 60)) # jak długo przebywa w pomieszczeniu, od 40 min długo, od 20 min średnio, do 20 min krótko
|
2022-05-11 17:32:28 +02:00
|
|
|
if osoba.kolumna > 21:
|
|
|
|
z.append(0) # zimne pomieszczenie
|
|
|
|
else:
|
|
|
|
z.append(1) # normalne pomieszczenie
|
|
|
|
z.extend(random.choices([1, 2], weights=[7, 3], k=1)) # poziom kurzu, 2 to największy/najbardziej niebezpieczny
|
|
|
|
z.extend(random.choices([1, 2], weights=[4, 6], k=1)) # poziom oświetlenia, 2 to najlepsze oświetlenie
|
|
|
|
if (0<=osoba.wiersz or osoba.wiersz<=13) and (17<=osoba.kolumna or osoba.kolumna<=19): #obok szafki z niebezpiecznymi towarami
|
|
|
|
z.append(1)
|
|
|
|
else:
|
|
|
|
z.append(0)
|
|
|
|
columns = ['plec', 'wiek', 'czas_w_pom', 'temp_w_pom', 'poziom_kurzu', 'poziom_oswietlenia', 'niebezp_towary']
|
|
|
|
z1 = pd.DataFrame([z],columns=columns)
|
|
|
|
z_pred = clf.predict(z1)
|
|
|
|
#print(z)
|
|
|
|
#print(z_pred)
|
|
|
|
return (z_pred)
|