Skip to content
pyAgrum
pypi downloads
pypi downloads
pypi downloads
pypi downloads

Using pyAgrum

Creative Commons LicenseaGrUMinteractive online version
import numpy as np
import matplotlib.pyplot as plt

Initialisation

Section titled “Initialisation”
  • importing pyAgrum
  • importing pyagrum.lib tools
  • loading a BN
import pyagrum as gum
import pyagrum.lib.notebook as gnb


gnb.configuration()
LibraryVersion
OSposix [darwin]
Python3.14.8 (main, Sep 30 2026, 17:55:09) [Clang 21.0.0 (clang-2100.3.34.2)]
IPython9.17.1
Matplotlib3.11.2
Numpy2.5.3
pyDot4.0.1
pyAgrum3.2.1
Fri Oct 09 08:52:22 2026 CEST
bn = gum.loadBN("res/alarm.dsl")
gnb.showBN(bn, size="9")

svg

Visualisation and inspection

Section titled “Visualisation and inspection”
print(bn["SHUNT"])
SHUNT:Labelized({NORMAL|HIGH})
print(bn.cpt(bn.idFromName("SHUNT")))
             ║  SHUNT            │
PULMEM│INTUBA║NORMAL   │HIGH     │
──────│──────║─────────│─────────│
TRUE  │NORMAL║ 0.1000  │ 0.9000  │
FALSE │NORMAL║ 0.9500  │ 0.0500  │
TRUE  │ESOPHA║ 0.1000  │ 0.9000  │
FALSE │ESOPHA║ 0.9500  │ 0.0500  │
TRUE  │ONESID║ 0.0100  │ 0.9900  │
FALSE │ONESID║ 0.0500  │ 0.9500  │
gnb.showTensor(bn.cpt(bn.idFromName("SHUNT")), digits=3)
SHUNT
INTUBATION
PULMEMBOLUS
NORMAL
HIGH
NORMAL
TRUE
0.1000.900
FALSE
0.9500.050
ESOPHAGEAL
TRUE
0.1000.900
FALSE
0.9500.050
ONESIDED
TRUE
0.0100.990
FALSE
0.0500.950

Results of inference

Section titled “Results of inference”

It is easy to look at result of inference

gnb.showPosterior(bn, {"SHUNT": "HIGH"}, "PRESS")
PyAgrum inline image 
gnb.showPosterior(bn, {"MINVOLSET": "NORMAL"}, "VENTALV")
PyAgrum inline image

Overall results

gnb.showInference(bn, size="10")

svg

What is the impact of observed variables (SHUNT and VENTALV for instance) on another on (PRESS) ?

ie = gum.LazyPropagation(bn)
ie.evidenceImpact("PRESS", ["SHUNT", "VENTALV"])
PRESS
SHUNT
VENTALV
ZERO
LOW
NORMAL
HIGH
NORMAL
ZERO
0.05690.26690.20050.4757
LOW
0.02080.25150.05530.6724
NORMAL
0.07690.32670.17720.4192
HIGH
0.05010.16330.27960.5071
HIGH
ZERO
0.05890.27260.19970.4688
LOW
0.03180.22370.05210.6924
NORMAL
0.17350.58390.14020.1024
HIGH
0.07110.23470.25330.4410

Using inference as a function

Section titled “Using inference as a function”

It is also easy to use inference as a routine in more complex procedures.

import time


r = range(0, 100)
xs = [x / 100.0 for x in r]


tf = time.time()
ys = [gum.getPosterior(bn, evs={"MINVOLSET": [0, x / 100.0, 0.5]}, target="VENTALV").tolist() for x in r]
delta = time.time() - tf


plt.plot(xs, ys)
plt.legend([bn["VENTALV"].label(i) for i in range(bn["VENTALV"].domainSize())], loc=7)
plt.title("VENTALV (100 inferences in %d ms)" % delta)
plt.ylabel("posterior Probability")
plt.xlabel("Evidence on MINVOLSET : [0,x,0.5]")
plt.show()

svg

Another example : python gives access to a large set of tools. Here the value for the equality of two probabilities of a posterior is easely computed.

x = [p / 100.0 for p in range(0, 100)]


tf = time.time()
y = [
  gum.getPosterior(bn, evs={"HRBP": [1.0 - p / 100.0, 1.0 - p / 100.0, p / 100.0]}, target="TPR").tolist()
  for p in range(0, 100)
]
delta = time.time() - tf


plt.plot(x, y)
plt.title("HRBP (100 inferences in %d ms)" % delta)
v = bn["TPR"]
plt.legend([v.label(i) for i in range(v.domainSize())], loc="best")
np1 = (np.transpose(y)[0] > np.transpose(y)[2]).argmin()
plt.text(x[np1] - 0.05, y[np1][0] + 0.005, str(x[np1]), bbox=dict(facecolor="red", alpha=0.05))
plt.show()

svg

BN as a classifier

Section titled “BN as a classifier”

Generation of databases

Section titled “Generation of databases”

Using the CSV format for the database:

print(
  f"The log2-likelihood of the generated base : {gum.generateSample(bn, 1000, 'out/test.csv', with_labels=True):.2f}"
)
The log2-likelihood of the generated base : -15000.71
with open("out/test.csv", "r") as src:
  for _ in range(10):
    print(src.readline(), end="")
ARTCO2,VENTMACH,ANAPHYLAXIS,TPR,LVEDVOLUME,HYPOVOLEMIA,SHUNT,DISCONNECT,PVSAT,BP,SAO2,HISTORY,MINVOLSET,VENTLUNG,HRBP,HREKG,CATECHOL,CVP,ERRLOWOUTPUT,PRESS,INSUFFANESTH,KINKEDTUBE,STROKEVOLUME,HR,LVFAILURE,CO,INTUBATION,FIO2,VENTTUBE,PAP,VENTALV,PULMEMBOLUS,EXPCO2,HRSAT,PCWP,MINVOL,ERRCAUTER
HIGH,NORMAL,FALSE,LOW,HIGH,TRUE,NORMAL,FALSE,LOW,LOW,LOW,FALSE,NORMAL,ZERO,LOW,HIGH,HIGH,HIGH,FALSE,HIGH,FALSE,FALSE,LOW,NORMAL,FALSE,LOW,NORMAL,NORMAL,LOW,NORMAL,ZERO,FALSE,LOW,LOW,HIGH,ZERO,FALSE
HIGH,NORMAL,FALSE,NORMAL,NORMAL,FALSE,NORMAL,FALSE,NORMAL,LOW,LOW,FALSE,NORMAL,HIGH,LOW,NORMAL,NORMAL,NORMAL,FALSE,HIGH,FALSE,TRUE,NORMAL,NORMAL,FALSE,NORMAL,NORMAL,NORMAL,LOW,NORMAL,LOW,FALSE,HIGH,LOW,NORMAL,LOW,TRUE
HIGH,NORMAL,FALSE,HIGH,NORMAL,FALSE,HIGH,FALSE,LOW,HIGH,LOW,FALSE,NORMAL,ZERO,HIGH,NORMAL,HIGH,NORMAL,FALSE,HIGH,FALSE,FALSE,NORMAL,HIGH,FALSE,HIGH,NORMAL,NORMAL,LOW,NORMAL,ZERO,FALSE,LOW,HIGH,NORMAL,ZERO,FALSE
HIGH,NORMAL,FALSE,NORMAL,HIGH,FALSE,NORMAL,FALSE,LOW,HIGH,LOW,FALSE,NORMAL,ZERO,HIGH,HIGH,HIGH,HIGH,FALSE,HIGH,FALSE,FALSE,NORMAL,HIGH,FALSE,HIGH,NORMAL,NORMAL,LOW,NORMAL,ZERO,FALSE,LOW,HIGH,HIGH,ZERO,FALSE
HIGH,NORMAL,FALSE,NORMAL,NORMAL,FALSE,NORMAL,FALSE,LOW,HIGH,LOW,FALSE,NORMAL,ZERO,HIGH,HIGH,HIGH,NORMAL,FALSE,HIGH,TRUE,FALSE,NORMAL,HIGH,FALSE,HIGH,NORMAL,NORMAL,LOW,NORMAL,ZERO,FALSE,LOW,HIGH,NORMAL,ZERO,FALSE
HIGH,NORMAL,FALSE,LOW,NORMAL,FALSE,NORMAL,FALSE,NORMAL,LOW,LOW,FALSE,NORMAL,HIGH,HIGH,HIGH,HIGH,NORMAL,FALSE,HIGH,FALSE,TRUE,NORMAL,HIGH,FALSE,HIGH,NORMAL,NORMAL,LOW,NORMAL,LOW,FALSE,HIGH,HIGH,NORMAL,ZERO,FALSE
HIGH,ZERO,FALSE,LOW,NORMAL,FALSE,NORMAL,FALSE,LOW,LOW,LOW,FALSE,NORMAL,ZERO,NORMAL,NORMAL,HIGH,HIGH,FALSE,ZERO,FALSE,FALSE,NORMAL,LOW,FALSE,LOW,NORMAL,NORMAL,ZERO,NORMAL,ZERO,FALSE,LOW,LOW,NORMAL,LOW,TRUE
HIGH,NORMAL,FALSE,NORMAL,NORMAL,FALSE,HIGH,FALSE,LOW,NORMAL,LOW,FALSE,NORMAL,ZERO,HIGH,HIGH,HIGH,LOW,FALSE,HIGH,FALSE,FALSE,NORMAL,HIGH,FALSE,NORMAL,NORMAL,NORMAL,LOW,NORMAL,ZERO,FALSE,LOW,HIGH,NORMAL,ZERO,FALSE
HIGH,NORMAL,FALSE,HIGH,NORMAL,FALSE,NORMAL,FALSE,LOW,HIGH,LOW,FALSE,NORMAL,ZERO,HIGH,HIGH,HIGH,NORMAL,FALSE,NORMAL,TRUE,FALSE,NORMAL,HIGH,FALSE,HIGH,NORMAL,NORMAL,LOW,LOW,ZERO,FALSE,LOW,HIGH,NORMAL,ZERO,FALSE

probabilistic classifier using BN

Section titled “probabilistic classifier using BN”

(because of the use of from-bn-generated csv files, quite good ROC curves are expected)

from pyagrum.lib.bn2roc import showROC_PR


showROC_PR(
  bn,
  "out/test.csv",
  target="CATECHOL",
  label="HIGH",  # class and label
  show_progress=True,
  show_fig=True,
  with_labels=True,
)

out/test.csv: 0%| |

out/test.csv: 100%|██████████|

svg

(np.float64(0.949145252214657),
 np.float64(0.9300899828),
 np.float64(0.9959010352129366),
 np.float64(0.4221808066))

Using another class variable

showROC_PR(bn, "out/test.csv", "SAO2", "HIGH", show_progress=True)

out/test.csv: 0%| |

out/test.csv: 100%|██████████|

svg

(np.float64(0.16797996011479158),
 np.float64(0.012257517612538345),
 np.float64(0.013142647998069768),
 np.float64(0.012257517612538345))

Fast prototyping for BNs

Section titled “Fast prototyping for BNs”
bn1 = gum.fastBN("a->b;a->c;b->c;c->d", 3)


gnb.sideBySide(
  *[gnb.getInference(bn1, evs={"c": val}, targets={"a", "c", "d"}) for val in range(3)],
  captions=[f"Inference given that $c={val}$" for val in range(3)],
)
structs Inference in   0.11ms a 2026-10-09T08:52:28.072795 image/svg+xml Matplotlib v3.11.2, b b a->b c 2026-10-09T08:52:28.151011 image/svg+xml Matplotlib v3.11.2, a->c b->c d 2026-10-09T08:52:28.176255 image/svg+xml Matplotlib v3.11.2, c->d
Inference given that $c=0$
structs Inference in   0.11ms a 2026-10-09T08:52:28.501439 image/svg+xml Matplotlib v3.11.2, b b a->b c 2026-10-09T08:52:28.535461 image/svg+xml Matplotlib v3.11.2, a->c b->c d 2026-10-09T08:52:28.589645 image/svg+xml Matplotlib v3.11.2, c->d
Inference given that $c=1$
structs Inference in   0.13ms a 2026-10-09T08:52:29.005009 image/svg+xml Matplotlib v3.11.2, b b a->b c 2026-10-09T08:52:29.060235 image/svg+xml Matplotlib v3.11.2, a->c b->c d 2026-10-09T08:52:29.095185 image/svg+xml Matplotlib v3.11.2, c->d
Inference given that $c=2$
print(gum.getPosterior(bn1, evs={"c": 0}, target="c"))
print(gum.getPosterior(bn1, evs={"c": 0}, target="d"))


## using pyagrum.lib.notebook's helpers
gnb.flow.row(gum.getPosterior(bn1, evs={"c": 0}, target="c"), gum.getPosterior(bn1, evs={"c": 0}, target="d"))
  c                          │
0        │1        │2        │
─────────│─────────│─────────│
 1.0000  │ 0.0000  │ 0.0000  │




  d                          │
0        │1        │2        │
─────────│─────────│─────────│
 0.6026  │ 0.0943  │ 0.3031  │
c
0
1
2
1.00000.00000.0000
d
0
1
2
0.60260.09430.3031

Joint posterior, impact of multiple evidence

Section titled “Joint posterior, impact of multiple evidence”
bn = gum.fastBN("a->b->c->d;b->e->d->f;g->c")
gnb.sideBySide(bn, gnb.getInference(bn))
G c c d d c->d b b b->c e e b->e a a a->b e->d g g g->c f f d->f
structs Inference in   0.43ms a 2026-10-09T08:52:29.439645 image/svg+xml Matplotlib v3.11.2, b 2026-10-09T08:52:29.488545 image/svg+xml Matplotlib v3.11.2, a->b c 2026-10-09T08:52:29.534932 image/svg+xml Matplotlib v3.11.2, b->c e 2026-10-09T08:52:29.608246 image/svg+xml Matplotlib v3.11.2, b->e d 2026-10-09T08:52:29.574056 image/svg+xml Matplotlib v3.11.2, c->d f 2026-10-09T08:52:29.648317 image/svg+xml Matplotlib v3.11.2, d->f e->d g 2026-10-09T08:52:29.682506 image/svg+xml Matplotlib v3.11.2, g->c 
ie = gum.LazyPropagation(bn)
ie.addJointTarget({"e", "f", "g"})
ie.makeInference()
gnb.sideBySide(
  ie.jointPosterior({"e", "f", "g"}),
  ie.jointPosterior({"e", "g"}),
  captions=["Joint posterior $P(e,f,g)$", "Joint posterior $P(e,f)$"],
)
g
f
e
0
1
0
0
0.26940.0601
1
0.09100.0200
1
0
0.34500.0753
1
0.11400.0252

Joint posterior $P(e,f,g)$
g
e
0
1
0
0.61440.1354
1
0.20500.0452

Joint posterior $P(e,f)$
gnb.sideBySide(
  ie.evidenceImpact("a", ["e", "f"]),
  ie.evidenceImpact("a", ["d", "e", "f"]),
  captions=["$\\forall e,f, P(a|e,f)$", "$\\forall d,e,f, P(a|d,e,f)=P(a|d,e)$ using d-separation"],
)
a
f
e
0
1
0
0
0.39280.6072
1
0.46890.5311
1
0
0.39140.6086
1
0.46830.5317

$\forall e,f, P(a|e,f)$
a
e
d
0
1
0
0
0.39360.6064
1
0.39110.6089
1
0
0.46930.5307
1
0.46810.5319

$\forall d,e,f, P(a|d,e,f)=P(a|d,e)$ using d-separation
gnb.sideBySide(
  ie.evidenceJointImpact(["a", "b"], ["e", "f"]),
  ie.evidenceJointImpact(["a", "b"], ["d", "e", "f"]),
  captions=["$\\forall e,f, P(a,b|e,f)$", "$\\forall d,e,f, P(a,b|d,e,f)=P(a,b|d,e)$ using d-separation"],
)
b
f
e
a
0
1
0
0
0
0.30800.0847
1
0.52120.0860
1
0
0.08130.3876
1
0.13760.3935
1
0
0
0.31210.0793
1
0.52800.0805
1
0
0.08330.3850
1
0.14090.3908

$\forall e,f, P(a,b|e,f)$
b
e
d
a
0
1
0
0
0
0.30550.0881
1
0.51690.0895
1
0
0.31290.0782
1
0.52950.0794
1
0
0
0.08010.3892
1
0.13550.3951
1
0
0.08370.3844
1
0.14160.3902

$\forall d,e,f, P(a,b|d,e,f)=P(a,b|d,e)$ using d-separation

Most Probable Explanation

Section titled “Most Probable Explanation”

The Most Probable Explanation (MPE) is a concept commonly used in the field of probabilistic reasoning and Bayesian statistics. It refers to the set of values for the variables in a given probabilistic model that is the most consistent with (that maximizes the likelihood of) the observed evidence. Essentially, it represents the most likely scenario or explanation given the available evidenceand the underlying probabilistic model.

ie = gum.LazyPropagation(bn)
print(ie.mpe())

<d:1|e:0|c:0|b:0|a:1|g:0|f:1>

evs = {"e": 0, "g": 0}
ie.setEvidence(evs)
vals = ie.mpeLog2Posterior()
print(
  f"The most probable explanation for observation {evs} is \n - the configuration {vals[0]} \n - for a log probability of {vals[1]:.6f}"
)
The most probable explanation for observation {'e': 0, 'g': 0} is
 - the configuration <g:0|e:0|d:1|f:1|c:0|b:0|a:1>
 - for a log probability of -2.602178