def pyfgs(d):
'''
fast greedy equivalence search from pycausal
https://github.com/bd2kccd/py-causal
Interpretation reasoning is provided in paper linked in repo README
'''
pc.start_vm()
n = d.shape[0]
d = np.asarray(np.r_[d[:, :-1], d[:, 1:]])
d = pd.DataFrame(d)
d = d.T
d.columns = [str(x + 1) + '*'
for x in range(n)] + [str(x + 1) for x in range(n)]
fgs = search.fges(d,
penaltydiscount=2,
faithfulnessAssumed=True,
verbose=True)
edg = fgs.edges
print(edg)
g2 = {}
for i in np.arange(1, n + 1):
g2[str(i)] = {}
for i in edg:
if re.match(r'[0-9]+\s.*[0-9]+\Z', i):
y = re.findall(r'[0-9]+', i)
if y[1] in g2[y[0]].keys():
g2[y[0]][y[1]].add((2, 0))
else:
g2[y[0]][y[1]] = set([(2, 0)])
if y[0] in g2[y[1]].keys():
g2[y[1]][y[0]].add((2, 0))
else:
g2[y[1]][y[0]] = set([(2, 0)])
elif re.match(r'[0-9]+\*.*[0-9]+\Z', i):
y = re.findall(r'[0-9]+', i)
if y[1] in g2[y[0]].keys():
g2[y[0]][y[1]].add((0, 1))
else:
g2[y[0]][y[1]] = set([(0, 1)])
elif re.match(r'[0-9]+\s.*[0-9]+\*', i):
y = re.findall(r'[0-9]+', i)
if y[0] in g2[y[1]].keys():
g2[y[1]][y[0]].add((0, 1))
else:
g2[y[1]][y[0]] = set([(0, 1)])
return conv.dict_format_converter(g2)
#!/usr/local/bin/python
import os
import pandas as pd
import pydot
from IPython.display import SVG
data_dir = os.path.join(os.getcwd(), 'data', 'audiology.txt')
df = pd.read_table(data_dir, sep="\t")
from pycausal import pycausal as pc
pc.start_vm(java_max_heap_size = '100M')
from pycausal import search as s
fges = s.fgesDiscrete(df,structurePrior = 1.0, samplePrior = 1.0,
maxDegree = 3, faithfulnessAssumed = True, numOfThreads = 2,
verbose = True)
fges.getNodes()
fges.getEdges()
dot = fges.getDot()
svg_str = dot.create_svg(prog='dot')
f = open('fges-discrete.dot','w')
f.write(svg_str)
f.close()
#!/usr/local/bin/python
import os
import pandas as pd
import pydot
from IPython.display import SVG
data_dir = os.path.join(os.getcwd(), 'data', 'charity.txt')
df = pd.read_table(data_dir, sep="\t")
from pycausal import pycausal as pc
pc.start_vm()
from pycausal import search as s
fges = s.fges(df,
penaltydiscount=2,
maxDegree=-1,
faithfulnessAssumed=True,
verbose=True)
fges.getNodes()
fges.getEdges()
dot = fges.getDot()
svg_str = dot.create_svg(prog='dot')
f = open('fges-continuous.dot', 'w')
f.write(svg_str)
f.close()
#!/usr/local/bin/python
import os
import pandas as pd
import pydot
from IPython.display import SVG
data_dir = os.path.join(os.getcwd(), 'data', 'audiology.txt')
df = pd.read_table(data_dir, sep="\t")
from pycausal import pycausal as pc
pc.start_vm(java_max_heap_size='100M')
from pycausal import search as s
fges = s.fgesDiscrete(df,
structurePrior=1.0,
samplePrior=1.0,
maxDegree=3,
faithfulnessAssumed=True,
verbose=True)
fges.getNodes()
fges.getEdges()
dot = fges.getDot()
svg_str = dot.create_svg(prog='dot')
f = open('fges-discrete.dot', 'w')
f.write(svg_str)
#!/usr/local/bin/python
import os
import pandas as pd
import pydot
from IPython.display import SVG
data_dir = os.path.join(os.getcwd(), 'data', 'charity.txt')
df = pd.read_table(data_dir, sep="\t")
from pycausal import pycausal as pc
pc.start_vm()
from pycausal import search as s
fges = s.fges(df,penaltydiscount = 2, maxDegree = -1,
faithfulnessAssumed = True, numofthreads = 2, verbose = True)
fges.getNodes()
fges.getEdges()
dot = fges.getDot()
svg_str = dot.create_svg(prog='dot')
f = open('fges-continuous.dot','w')
f.write(svg_str)
f.close()
pc.stop_vm()