def apply_article_metadata(article = None):
if article:
artObs = [article]
num_arts = 1
else:
# artObs = m.Article.objects.filter(metadata__isnull = True, articlefulltext__isnull = False).distinct()
artObs = m.Article.objects.filter(articlefulltext__isnull = False).distinct()
# artObs = artObs.exclude(articlefulltext__articlefulltextstat__metadata_processed = True)
# artObs = m.Article.objects.filter(articlefulltext__isnull = False, articlefulltext__articlefulltextstat__methods_tag_found = True).distinct()
# artObs = artObs.exclude(articlefulltext__articlefulltextstat__metadata_processed = True)
# artObs = artObs.exclude(articlefulltext__articlefulltextstat__metadata_human_assigned = True)
num_arts = artObs.count()
print 'annotating %s articles for metadata...' % num_arts
for i,art in enumerate(artObs):
if not article:
prog(i,num_arts)
assign_metadata.assign_species(art)
assign_metadata.assign_electrode_type(art)
assign_metadata.assign_strain(art)
assign_metadata.assign_rec_temp(art)
assign_metadata.assign_prep_type(art)
assign_metadata.assign_animal_age(art)
assign_metadata.assign_jxn_potential(art)
assign_metadata.assign_solution_concs(art)
aft_ob = art.get_full_text()
aftStatOb = m.ArticleFullTextStat.objects.get_or_create(article_full_text = aft_ob)[0]
aftStatOb.metadata_processed = True
aftStatOb.save()
def loadEdges():
with open('neurotree/data/dumpedges.php','rU') as f:
reader = csv.reader(f)
junk = reader.next() # Skip headers .
junk = reader.next() # Skip headers .
rows = list(reader)
n_rows = len(rows)
with transaction.commit_on_success():
for i,row in enumerate(rows):
prog(i,n_rows)
try:
node1 = m.Node.objects.get(id=int(row[0]))
node2 = m.Node.objects.get(id=int(row[2]))
except:
print "One of these nodes not found: %d,%d" % \
(int(row[0]),int(row[2]))
pass
# Nodes probably did not exist.
# There are some edges in the file with node id's
# that are not in the dumpnodes.php file.
else:
edge,created = m.Edge.objects.get_or_create(
node1=node1,
node2=node2,
defaults={'relationcode':row[1],
'relationstring':row[3],
'stopyear':int(row[4])})
def assign_neuroner_ids():
"""Iterate through all neuron concept maps and assign neuroner ontology ids"""
ncms = m.NeuronConceptMap.objects.filter(times_validated__gt = 0)
# actually update objects
cm_expert_list_len = len(ncms)
for i,cm in enumerate(ncms):
prog(i,cm_expert_list_len)
if cm.neuron_long_name:
neuron_pref_name = cm.neuron_long_name
else:
neuron_pref_name = cm.neuron.name
neuroner_ids = get_neuroner_annotations(neuron_pref_name)
json_neuroner_ids = json.dumps(neuroner_ids)
# if annotations are unchanged, then don't save and continue
# TODO: fix the below, doesn't seem to work
if set(neuroner_ids) == set(cm.get_neuroner()):
#print 'existing neuroner ids, moving on'
continue
print 'adding new neuroner id to pk %s: ' % cm.pk
cm.neuroner_ids = json_neuroner_ids
# don't actually update the changed date
hcm = cm.history.all()[0]
cm._history_date = hcm.history_date
cm.save(new_history_time = True)
def graphviz_dot_plus_grandparents(author_list_full,
authors,
max_path_length=2):
"""After running this, run:
dot -Tpdf -Gcharset=latin1 authors.dot -o authors.pdf
on the command line."""
f = open('authors.dot','w')
f.write('digraph G {\r\n')
for i,author1 in enumerate(author_list_full):
prog(i,len(author_list_full))
f.write('\t%s\r\n' % clean(author1.lastname))
if author1 in authors:
f.write('\t%s\r[fillcolor = red]\r\n' % clean(author1.lastname))
for author2 in author_list_full:
if author1 == author2:
continue
path = shortest_path(author1,
author2,
directions=['up'],
max_path_length=max_path_length)
if path is not None:
#f.write('\t%s->%s\r\n' % (clean(author1.lastname),clean(author2.lastname)))
for j in range(len(path)-1):
f.write('\t%s->%s\r\n' % \
(clean(path[j].lastname),clean(path[j+1].lastname)))
print path
f.write('}')
f.close()
def assign_stat_object_to_data_tables():
"""go through each data table object and add info about who curated and when"""
dts = m.DataTable.objects.filter(datasource__ephysconceptmap__isnull=False)
num_dts = dts.count()
for i, dt in enumerate(dts):
print "updating data tables with summary objects"
prog(i, num_dts)
update_data_table_stat(dt)
def author_path_length_matrix(authors): import neurotree.neurotree_author_search as search import numpy as np #authors = search.get_neurotree_authors()[0] matrix = np.zeros((len(authors),len(authors))) for i in range(len(authors)): prog(i,len(authors)) for j in range(len(authors)): path = shortest_path(authors[i],authors[j], directions=['up'],max_path_length=3) matrix[i,j] = len(path) if path is not None else None return matrix
def get_author_grandparents(authors):
"""
Returns a list of neurotree nodes corresponding to academic
grandparents of input neurotree nodes
"""
author_list_initial = authors
num_authors = len(authors)
author_list_final = []
relationcodes = [1, 2]
for i,a in enumerate(author_list_initial):
prog(i,num_authors)
a_relations = a.parents.filter(relationcode__in=relationcodes)
parent_nodes = [x.node2 for x in a_relations]
#print parent_nodes
author_list_final.extend(parent_nodes)
for p in parent_nodes:
p_relations = p.parents.filter(relationcode__in=relationcodes)
grand_parent_nodes = [y.node2 for y in p_relations]
author_list_final.extend(grand_parent_nodes)
author_list_final.extend(author_list_initial)
return list(set(author_list_final))
def loadNodes():
with open('neurotree/data/dumpnodes.php','rU') as f:
reader = csv.reader(f)
junk = reader.next() # Skip leading blank line.
junk = reader.next() # Skip header.
rows = list(reader)
n_rows = len(rows)
with transaction.commit_on_success():
for i,row in enumerate(rows):
#if float(i)/n_rows < 0.6:
# continue
prog(i,n_rows)
if len(row)<7: # Probably the last line (blank).
continue
trees = []
tree_names = row[6].split('+')
for tree_name in tree_names:
tree,created = m.Tree.objects.get_or_create(name=tree_name)
trees.append(tree)
firstname = row[1].decode('latin1')
middlename = row[2].decode('latin1')
lastname = row[3].decode('latin1')
location = row[4].decode('latin1')
altpubmed = row[5].decode('latin1')
#global row_
#row_ = row
#print firstname,middlename,lastname,location
node,created = m.Node.objects.get_or_create(id=row[0],
defaults={'firstname':firstname,
'middlename':middlename,
'lastname':lastname,
'location':location,
'altpubmed':altpubmed})
for tree in trees:
node.tree.add(tree)
