def unlink_concept_name(self, cui, name, full_unlink=True):
names = [name, name.lower()]
# Unlink a concept from a name
p_name, tokens, _, _ = get_all_from_name(name=name, source_value=name, nlp=self.nlp, version='clean')
# Add the clean version of the name
names.append(p_name)
# Get the raw version
p_name, tokens, _, _ = get_all_from_name(name=name, source_value=name, nlp=self.nlp, version='raw')
# Append the raw evrsion
names.append(p_name)
if tokens[-1].lower() == "s":
# Remove last 's' - a stupid bug
names.append(p_name[0:-1])
for name in names:
cuis = [cui]
if full_unlink and name in self.cdb.name2cui:
cuis = list(self.cdb.name2cui[name])
for cui in cuis:
if cui in self.cdb.cui2names and name in self.cdb.cui2names[cui]:
self.cdb.cui2names[cui].remove(name)
if len(self.cdb.cui2names[cui]) == 0:
del self.cdb.cui2names[cui]
if name in self.cdb.name2cui:
if cui in self.cdb.name2cui[name]:
self.cdb.name2cui[name].remove(cui)
if len(self.cdb.name2cui[name]) == 0:
del self.cdb.name2cui[name]
def unlink_concept_name(self, cui, name, full_unlink=True):
r'''
Unlink a concept name from the CUI (or all CUIs if full_unlink), removes the link from
the Concept Database (CDB). As a consequence medcat will never again link the `name`
to this CUI - meaning the name will not be detected as a concept in the future.
Args:
cui (str):
The CUI from which the `name` will be removed
name (str):
The span of text to be removed from the linking dictionary
full_unlink (boolean):
If True, the `name` will not only be removed from the given `cui` but from
each concept in the database that is associated with this name.
Examples:
>>> # To never again link C0020538 to HTN
>>> cat.unlink_concept_name('C0020538', 'htn', False)
'''
names = [name, name.lower()]
# Unlink a concept from a name
p_name, tokens, _, _ = get_all_from_name(name=name,
source_value=name,
nlp=self.nlp,
version='clean')
# Add the clean version of the name
names.append(p_name)
# Get the raw version
p_name, tokens, _, _ = get_all_from_name(name=name,
source_value=name,
nlp=self.nlp,
version='raw')
# Append the raw evrsion
names.append(p_name)
if tokens[-1].lower() == "s":
# Remove last 's' - a stupid bug
names.append(p_name[0:-1])
for name in names:
cuis = [cui]
if full_unlink and name in self.cdb.name2cui:
cuis = list(self.cdb.name2cui[name])
for cui in cuis:
if cui in self.cdb.cui2names and name in self.cdb.cui2names[
cui]:
self.cdb.cui2names[cui].remove(name)
if len(self.cdb.cui2names[cui]) == 0:
del self.cdb.cui2names[cui]
if name in self.cdb.name2cui:
if cui in self.cdb.name2cui[name]:
self.cdb.name2cui[name].remove(cui)
if len(self.cdb.name2cui[name]) == 0:
del self.cdb.name2cui[name]
def add_concept(self, concept, text=None, tkn_inds=None):
cui = concept['cui']
onto = concept.get('onto', 'user')
pretty_name = concept['name']
source_value = concept['source_value']
name, tokens, snames, tokens_vocab = get_all_from_name(
name=pretty_name, source_value=source_value, nlp=self.nlp)
tui = concept.get('tui', 'None')
unique = True
# Add the new concept
self.cdb.add_concept(cui,
name,
onto,
tokens,
snames,
tui=tui,
pretty_name=pretty_name,
is_pref_name=True,
tokens_vocab=tokens_vocab,
unique=unique,
original_name=source_value)
if tkn_inds and text:
# Add the context
self.add_concept_cntx(cui, text, tkn_inds)
def _add_name(self, cui, source_val, is_pref_name, only_new=False, desc=None, tui=None):
onto = 'def'
all_cuis = []
if cui in self.cdb.cui2ontos and self.cdb.cui2ontos[cui]:
onto = list(self.cdb.cui2ontos[cui])[0]
# Add the original version of the name just lowercased
p_name, tokens, snames, tokens_vocab = get_all_from_name(name=source_val,
source_value=source_val,
nlp=self.nlp, version='none')
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[cui]:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui, p_name, onto, tokens, snames, tokens_vocab=tokens_vocab,
original_name=source_val, is_pref_name=False, desc=desc, tui=tui)
all_cuis.extend(self.cdb.name2cui[p_name])
p_name, tokens, snames, tokens_vocab = get_all_from_name(name=source_val,
source_value=source_val,
nlp=self.nlp, version='clean')
# This will add a new concept if the cui doesn't exist
# or link the name to an existing concept if it exists.
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[cui]:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui, p_name, onto, tokens, snames, tokens_vocab=tokens_vocab,
original_name=source_val, is_pref_name=False, desc=desc, tui=tui)
all_cuis.extend(self.cdb.name2cui[p_name])
# Add the raw also if needed
p_name, tokens, snames, tokens_vocab = get_all_from_name(name=source_val,
source_value=source_val,
nlp=self.nlp, version='raw')
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[cui] or is_pref_name:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui, p_name, onto, tokens, snames, tokens_vocab=tokens_vocab,
original_name=source_val, is_pref_name=is_pref_name, desc=desc, tui=tui)
all_cuis.extend(self.cdb.name2cui[p_name])
# Fix for ntkns in cdb
if p_name in self.cdb.name2ntkns:
if len(tokens) not in self.cdb.name2ntkns[p_name]:
self.cdb.name2ntkns[p_name].add(len(tokens))
return list(set(all_cuis))
def _add_name(self, cui, source_val, is_pref_name, only_new=False):
onto = 'def'
if cui in self.cdb.cui2ontos and self.cdb.cui2ontos[cui]:
onto = list(self.cdb.cui2ontos[cui])[0]
p_name, tokens, snames, tokens_vocab = get_all_from_name(name=source_val,
source_value=source_val,
nlp=self.nlp, version='clean')
# This will add a new concept if the cui doesn't exist
#or link the name to an existing concept if it exists.
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[cui]:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui, p_name, onto, tokens, snames, tokens_vocab=tokens_vocab,
original_name=source_val, is_pref_name=False)
# Add the raw also if needed
p_name, tokens, snames, tokens_vocab = get_all_from_name(name=source_val,
source_value=source_val,
nlp=self.nlp, version='raw')
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[cui] or is_pref_name:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui, p_name, onto, tokens, snames, tokens_vocab=tokens_vocab,
original_name=source_val, is_pref_name=is_pref_name)
def add_name(self, cui, name):
onto = 'def'
if cui in self.cdb.cui2onto:
onto = self.cdb.cui2onto[cui][0]
p_name, tokens, snames, tokens_vocab = get_all_from_name(
name=name, source_value=name, nlp=self.nlp)
# This will add a new concept if the cui doesn't exist
#or link the name to an existing concept if it exists.
self.cdb.add_concept(cui,
p_name,
onto,
tokens,
snames,
tokens_vocab=tokens_vocab,
original_name=name)
def unlink_concept_name(self, cui, name):
# Unlink a concept from a name
p_name, _, _, _ = get_all_from_name(name=name, source_value=name, nlp=self.nlp)
# To be sure unlink the orignal and the processed name
names = [name, p_name]
for name in names:
if name in self.cdb.cui2names[cui]:
self.cdb.cui2names[cui].remove(name)
if len(self.cdb.cui2names[cui]) == 0:
del self.cdb.cui2names[cui]
if name in self.cdb.name2cui:
if cui in self.cdb.name2cui[name]:
self.cdb.name2cui[name].remove(cui)
if len(self.cdb.name2cui[name]) == 0:
del self.cdb.name2cui[name]
def _add_name(self,
cui,
source_val,
is_pref_name,
only_new=False,
desc=None,
tui=None):
r'''
Please do not use directly. This function will add a name to a CUI (existing or new).
Args:
cui (str):
The CUI to which to add the name
source_val (str):
The `name` or span or source_value that will be linked to the cui
is_pref_name (boolean):
Is this source_val the prefered `name` for this CUI (concept)
only_new (bool):
Only add the name if it does not exist in the current CDB and is not linked
to any concept (CUI) in the current CDB.
desc (str):
Description for this concept
tui (str):
Semenantic Type identifer for this concept, should be a TUI that exisit in the
current CDB. Have a look at cdb.tui2names - for a list of all existing TUIs
in the current CDB.
Examples:
Do not use.
'''
onto = 'def'
all_cuis = []
if cui in self.cdb.cui2ontos and self.cdb.cui2ontos[cui]:
onto = list(self.cdb.cui2ontos[cui])[0]
# Add the original version of the name just lowercased
p_name, tokens, snames, tokens_vocab = get_all_from_name(
name=source_val,
source_value=source_val,
nlp=self.nlp,
version='none')
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[
cui]:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui,
p_name,
onto,
tokens,
snames,
tokens_vocab=tokens_vocab,
original_name=source_val,
is_pref_name=False,
desc=desc,
tui=tui)
all_cuis.extend(self.cdb.name2cui[p_name])
p_name, tokens, snames, tokens_vocab = get_all_from_name(
name=source_val,
source_value=source_val,
nlp=self.nlp,
version='clean')
# This will add a new concept if the cui doesn't exist
# or link the name to an existing concept if it exists.
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[
cui]:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui,
p_name,
onto,
tokens,
snames,
tokens_vocab=tokens_vocab,
original_name=source_val,
is_pref_name=False,
desc=desc,
tui=tui)
all_cuis.extend(self.cdb.name2cui[p_name])
# Add the raw also if needed
p_name, tokens, snames, tokens_vocab = get_all_from_name(
name=source_val,
source_value=source_val,
nlp=self.nlp,
version='raw')
if cui not in self.cdb.cui2names or p_name not in self.cdb.cui2names[
cui] or is_pref_name:
if not only_new or p_name not in self.cdb.name2cui:
self.cdb.add_concept(cui,
p_name,
onto,
tokens,
snames,
tokens_vocab=tokens_vocab,
original_name=source_val,
is_pref_name=is_pref_name,
desc=desc,
tui=tui)
all_cuis.extend(self.cdb.name2cui[p_name])
# Fix for ntkns in cdb
if p_name in self.cdb.name2ntkns:
if len(tokens) not in self.cdb.name2ntkns[p_name]:
self.cdb.name2ntkns[p_name].add(len(tokens))
return list(set(all_cuis))