from gPy.LearningUtils import * from gPy.Examples import asia from gPy.K2 import * from gPy.Models import CBN # generate 500 samples of the Asia network data = CausalWorld(asia) data.observe(5000) # for comparison true_model = asia true_adg = asia.adg() # search for the ADG found_adg = K2(max_parents_family=3).search(data, true_adg.topological_order()) # fit a model to the data using the ADG found_model = CBN.from_adg_data(found_adg, data) print 'True ADG\n', true_adg print 'Found ADG\n', found_adg print 'Structural Hamming distance:',shd(found_adg,true_adg) print 'BDeu scores of fitted models of' print 'found ADG:', bdeu(found_adg,data) print ' true ADG:', bdeu(true_adg,data) print 'KL-divergence of fitted found model from true model:',dkl(true_model,found_model)
from gPy.Examples import asia
from gPy.Parameters import CompactFactor
from gPy.IO import read_csv
import sys
data = CompactFactor(read_csv(open(sys.argv[1])))
print asia.bdeu_score(data)
def score_adg(adg,data):
print '^^^^'
for child in adg.vertices():
parents = adg.parents(child)
family = parents | set([child])
data_cpt = data.makeFactor(family).makeCPT(child,False)
print child, data_cpt.bdeu_score()
print 'vvvvv'
print
adg = asia.adg()
score_adg(adg,data)
adg.remove_arrow('Smoking','Cancer')
score_adg(adg,data)
from gPy.Examples import asia
from gPy.Structures import UGraph
from Tkinter import Tk
root = Tk()
root.title("Maximum cardinality search demonstration")
asia.adg().moralise().gui_maximum_cardinality_search(root)
UGraph(range(1,11),((1,2),(1,3),(2,3),(2,10),(3,10),(4,5),
(4,7),(5,6),(5,9),(5,7),(6,7),(6,9),
(7,8),(7,9),(8,9),(8,10),(9,10))).gui_maximum_cardinality_search(root)
root.mainloop()
from gPy.LearningUtils import * from gPy.Examples import asia from gPy.PC import * from gPy.Models import CBN # generate 500 samples of the Asia network data = CausalWorld(asia) data.observe(500) # for comparison true_model = asia true_adg = asia.adg() # search for the ADG ## the PC algorithm consists of a search for independence relations using a ## test for conditional independence. In this case, the G-squared statistic is ## taken to be distributed as Chi-squared (this is only true asymptotically). ci = PCCI(G2Separator(data)) ## the IC algorithm of Verma & Pearl is used to resolve these conditional ## independencies into a PDAG. Note that this may fail if the found conditional ## independencies are inconsistent with an ADG! pdag = ICPattern(ci) ## Meek's algorithm is used to resolve this PDAG into a particular ADG found_adg = pdag.orient() ## or for short:
from gPy.Examples import asia
from Tkinter import Tk
from gPy.Utils import scrolled_frame
root=Tk()
top=scrolled_frame(root,yscroll=500, height=4000)
asia.gui_display(top)
asia.adg().gui_display(top)
for factor in asia:
factor *= 1
asia.gui_display(top)
asia.adg().moralise().gui_display(top)
root.mainloop()
from gPy.Examples import asia
from gPy.Structures import GraphicalReducedHyperGraph
from Tkinter import Tk
hg = asia.hypergraph()
hg.red()
hg = GraphicalReducedHyperGraph(hg)
root=Tk()
gr = asia.adg().moralise()
gr.add_line('Cancer','Bronchitis')
hg2 = gr.hypergraph()
hg2.gui_grahams_algorithm(root)
root.mainloop()
def setUp(self):
self._asia_pdag = asia.adg().essential_graph()
self._tetrad_pdag = Graph(vertices=self._asia_pdag.vertices()
,arrows=[('Tuberculosis','TbOrCa'),('Cancer','TbOrCa')]
,lines=[('Smoking','Cancer'),('Smoking','Bronchitis')
,('Bronchitis','Dyspnea')])
from gPy.Examples import asia
from gPy.Data import CompactFactor
from gPy.IO import read_csv
import sys
data = CompactFactor(read_csv(open(sys.argv[1])))
print asia.bdeu_score(data)
def score_adg(adg,data):
print '^^^^'
for child in adg.vertices():
parents = adg.parents(child)
family = parents | set([child])
data_cpt = data.makeFactor(family).makeCPT(child,False)
print child, data_cpt.bdeu_score()
print 'vvvvv'
print
adg = asia.adg()
score_adg(adg,data)
from gPy.Examples import asia
from Tkinter import *
root = Tk()
orders =(['VisitAsia', 'Tuberculosis', 'XRay', 'Dyspnea', 'Bronchitis',
'Smoking', 'TbOrCa','Cancer'],
['TbOrCa', 'VisitAsia', 'Tuberculosis', 'XRay', 'Bronchitis',
'Smoking', 'Dyspnea','Cancer'])
ig = asia.adg().moralise()
for order in orders:
ig.copy().gui_triangulate(root,order)
root.mainloop()
def testAsia(self):
from gPy.Examples import asia
indep = GraphCI(asia.adg(), True, True).independences()
for k in indep.keys():
print k, ":", indep[k]
import unittest
from agm2 import find_minimal_separator
from gPy.Examples import asia
asia_adg = asia.adg()
cases = (
('VisitAsia','Tuberculosis',None),
('VisitAsia','Smoking',frozenset()),
('Bronchitis','Cancer',frozenset(['Smoking']))
)
cases2 = (
('Smoking','Dyspnea',2),
('VisitAsia','XRay',1),
('VisitAsia','Dyspnea',1),
)
class Testagm2(unittest.TestCase):
def test_asia(self):
for case in cases:
self.assertEqual(
find_minimal_separator(asia_adg,case[0],case[1]),
case[2])
def test_asia2(self):
for case in cases2:
sep = find_minimal_separator(asia_adg,case[0],case[1])
self.assertEqual(len(sep),case[2])
import unittest
from agm2 import find_minimal_separator
from gPy.Examples import asia
asia_adg = asia.adg()
cases = (('VisitAsia', 'Tuberculosis', None),
('VisitAsia', 'Smoking', frozenset()), ('Bronchitis', 'Cancer',
frozenset(['Smoking'])))
cases2 = (
('Smoking', 'Dyspnea', 2),
('VisitAsia', 'XRay', 1),
('VisitAsia', 'Dyspnea', 1),
)
class Testagm2(unittest.TestCase):
def test_asia(self):
for case in cases:
self.assertEqual(
find_minimal_separator(asia_adg, case[0], case[1]), case[2])
def test_asia2(self):
for case in cases2:
sep = find_minimal_separator(asia_adg, case[0], case[1])
self.assertEqual(len(sep), case[2])
suite = unittest.makeSuite(Testagm2)
def testAsia(self):
from gPy.Examples import asia
indep = GraphCI(asia.adg(), True, True).independences()
for k in indep.keys():
print k, ':', indep[k]
from gPy.Examples import asia
from Tkinter import Tk
from gPy.Utils import scrolled_frame
root=Tk()
top=scrolled_frame(root,yscroll=5000, height=4000)
asia.gui_display(top)
asia.adg().gui_display(top,width=400,height=300)
for factor in asia:
factor *= 1
asia.gui_display(top)
asia.adg().moralise().gui_display(top,width=400,height=300)
root.mainloop()
from gPy.Examples import asia
from Tkinter import Tk
from gPy.Utils import scrolled_frame
root = Tk()
top = scrolled_frame(root, yscroll=5000, height=4000)
asia.gui_display(top)
asia.adg().gui_display(top, width=400, height=300)
for factor in asia:
factor *= 1
asia.gui_display(top)
asia.adg().moralise().gui_display(top, width=400, height=300)
root.mainloop()
from gPy.Examples import asia
from gPy.Structures import UGraph
from Tkinter import Tk
root = Tk()
root.title("Maximum cardinality search demonstration")
asia.adg().moralise().gui_maximum_cardinality_search(root)
UGraph(range(1, 11),
((1, 2), (1, 3), (2, 3), (2, 10), (3, 10), (4, 5), (4, 7), (5, 6),
(5, 9), (5, 7), (6, 7), (6, 9), (7, 8), (7, 9), (8, 9), (8, 10),
(9, 10))).gui_maximum_cardinality_search(root)
root.mainloop()
