pyAgrum-nightly 2.3.1.9.dev202512261765915415__cp310-abi3-macosx_10_15_x86_64.whl

pyagrum/lib/bn2scores.py

@@ -0,0 +1,217 @@
+############################################################################
+#   This file is part of the aGrUM/pyAgrum library.                        #
+#                                                                          #
+#   Copyright (c) 2005-2025 by                                             #
+#       - Pierre-Henri WUILLEMIN(_at_LIP6)                                 #
+#       - Christophe GONZALES(_at_AMU)                                     #
+#                                                                          #
+#   The aGrUM/pyAgrum library is free software; you can redistribute it    #
+#   and/or modify it under the terms of either :                           #
+#                                                                          #
+#    - the GNU Lesser General Public License as published by               #
+#      the Free Software Foundation, either version 3 of the License,      #
+#      or (at your option) any later version,                              #
+#    - the MIT license (MIT),                                              #
+#    - or both in dual license, as here.                                   #
+#                                                                          #
+#   (see https://agrum.gitlab.io/articles/dual-licenses-lgplv3mit.html)    #
+#                                                                          #
+#   This aGrUM/pyAgrum library is distributed in the hope that it will be  #
+#   useful, but WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,          #
+#   INCLUDING BUT NOT LIMITED TO THE WARRANTIES MERCHANTABILITY or FITNESS #
+#   FOR A PARTICULAR PURPOSE  AND NONINFRINGEMENT. IN NO EVENT SHALL THE   #
+#   AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER #
+#   LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,        #
+#   ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR  #
+#   OTHER DEALINGS IN THE SOFTWARE.                                        #
+#                                                                          #
+#   See LICENCES for more details.                                         #
+#                                                                          #
+#   SPDX-FileCopyrightText: Copyright 2005-2025                            #
+#       - Pierre-Henri WUILLEMIN(_at_LIP6)                                 #
+#       - Christophe GONZALES(_at_AMU)                                     #
+#   SPDX-License-Identifier: LGPL-3.0-or-later OR MIT                      #
+#                                                                          #
+#   Contact  : info_at_agrum_dot_org                                       #
+#   homepage : http://agrum.gitlab.io                                      #
+#   gitlab   : https://gitlab.com/agrumery/agrum                           #
+#                                                                          #
+############################################################################
+
+"""
+The purpose of this module is to provide tools for computing different scores from a BN.
+"""
+
+import csv
+import math
+
+import pyagrum as gum
+
+
+def lines_count(filename):
+  """count lines in a file"""
+  with open(filename) as f:
+    count = sum(1 for _ in f)
+
+  return count
+
+
+def checkCompatibility(bn, fields, csv_name):
+  """
+  check if the variables of the bn are in the fields
+
+  Parameters
+  ----------
+  bn: gum.BayesNet
+    the model
+  fields: Dict[str,int]
+    Dict of name,position in the file
+  csv_name: str
+    name of the csv file
+
+  Raises
+  ------
+  gum.DatabaseError
+    if a BN variable is not in fields
+
+  Returns
+  -------
+  Dict[int,str]
+    return a dictionary of position for BN variables in fields
+  """
+  res = {}
+  for field in bn.names():
+    if field not in fields:
+      raise gum.DatabaseError(f"** At least, field '{field}' is missing in {csv_name}")
+    res[bn.idFromName(field)] = fields[field]
+
+  return res
+
+
+def computeScores(bn_name, csv_name, visible=False, dialect=None):
+  """
+  Compute scores (likelihood, aic, bic, mdl, etc.) from a bn w.r.t to a csv
+
+  Parameters
+  ----------
+  bn_name : pyagrum.BayesNet | str
+    a gum.BayesianNetwork or a filename for a BN
+  csv_name : str
+     a filename for the CSV database
+  visible: bool
+    do we show the progress
+  dialect: csv.Dialect
+    if not provided, dialect will be inferred using csv.Sniffer().sniff(csvfile.read(1024))
+
+  Returns
+  -------
+  Tuple[float,Dict[str,float]]
+    percentDatabaseUsed,scores
+  """
+  if isinstance(bn_name, str):
+    bn = gum.loadBN(bn_name)
+  else:
+    bn = bn_name
+
+  nbr_lines = lines_count(csv_name) - 1
+
+  if dialect is None:
+    with open(csv_name, "r") as csvfile:
+      k = 1024
+      buffer = ""
+      while True:
+        try:
+          buffer += csvfile.read(k)
+          dialect = csv.Sniffer().sniff(buffer)
+          break
+        except csv.Error:
+          if k == -1:
+            # we already tried with all the file
+            raise gum.DatabaseError("csv.Sniffer could not determine delimiter even with the whole file")
+          if len(buffer) > 16384:
+            # we try with all the file now
+            k = -1
+
+  nbr_insignificant = 0
+  num_ligne = 1
+  likelihood = 0.0
+
+  with open(csv_name, "r") as csvfile:
+    batchReader = csv.reader(csvfile, dialect)
+
+    titre = next(batchReader)
+    fields = {}
+    for i, nom in enumerate(titre):
+      fields[nom] = i
+
+    positions = checkCompatibility(bn, fields, csv_name)
+
+    inst = bn.completeInstantiation()
+
+    if visible:
+      # tqdm is optional
+      # pylint: disable=import-outside-toplevel
+      from tqdm import tqdm
+
+      pbar = tqdm(total=nbr_lines, desc=csv_name, bar_format="{desc}: {percentage:3.0f}%|{bar}|")
+
+    for data in batchReader:
+      num_ligne += 1
+
+      for i in positions.keys():
+        try:
+          name = bn.variable(i).name()
+          inst.chgVal(name, _getIdLabel(inst, name, data[positions[i]]))
+        except gum.OutOfBounds:
+          print(
+            f"Out of bounds for ({i},{positions[i]}) : unknown id or label '{data[positions[i]]}' for the variable {inst.variable(i)}"
+          )
+
+      p = bn.jointProbability(inst)
+      if p == 0.0:
+        print(str(num_ligne) + ":" + str(inst))
+        nbr_insignificant += 1
+      else:
+        likelihood += math.log(p, 2)
+      if visible:
+        pbar.update()
+
+    if visible:
+      pbar.close()
+
+    nbr_arcs = 1.0 * bn.sizeArcs()
+    dim = 1.0 * bn.dim()
+
+    aic = likelihood - dim
+    aicc = 2 * aic - 2 * dim * (dim + 1) / (nbr_lines - dim + 1) if (nbr_lines - dim + 1 > 0) else "undefined"
+    bic = likelihood - dim * math.log(nbr_lines, 2)
+    mdl = likelihood - nbr_arcs * math.log(nbr_lines, 2) - 32 * dim  # 32=nbr bits for a params
+
+  return (
+    (nbr_lines - nbr_insignificant) * 100.0 / nbr_lines,
+    {"likelihood": likelihood, "aic": aic, "aicc": aicc, "bic": bic, "mdl": mdl},
+  )
+
+
+def _getIdLabel(inst, i, label):
+  """
+  Return the idLabel. if label is an int, it already is an idLabel
+
+  Parameters
+  ----------
+  inst : gum.Instantiation
+    the inst to find the idLabel
+  i: int
+    the id to find the variable
+  label: int|str
+    the label or id of label
+
+  Returns
+  -------
+  int
+    the id of label
+  """
+  if isinstance(label, int):  # an idLabel
+    return label
+
+  return inst.variable(i)[label]