def simple_comparison(list1,list2):
clist1 = load_conceptlist(
'../concepticondata/conceptlists/'+list1+'.tsv'
)
clist2 = load_conceptlist(
'../concepticondata/conceptlists/'+list2+'.tsv'
)
gidx1 = 'ENGLISH' if 'ENGLISH' in clist1['header'] else 'GLOSS'
gidx2 = 'ENGLISH' if 'ENGLISH' in clist2['header'] else 'GLOSS'
clist2_keys = []
for k in clist2:
if k not in ['header','splits','mergers']:
clist2_keys += [(clist2[k]['CONCEPTICON_ID'], k)]
clist2_keys = dict(clist2_keys)
coverage = []
count = 0
for k in clist1:
if k not in ['header', 'splits', 'mergers']:
cid = clist1[k]['CONCEPTICON_ID']
if cid in clist2_keys:
coverage += [(cid, k, clist1[k][gidx1],
clist2_keys[cid], clist2[clist2_keys[cid]][gidx2])]
count += 1
else:
coverage += [(cid, k, clist1[k][gidx1],'?','?')]
for line in coverage:
print('\t'.join(line))
print(count / len(clist1), count)
def simple_comparison(list1,list2):
clist1 = load_conceptlist(
'../concepticondata/conceptlists/'+list1+'.tsv'
)
clist2 = load_conceptlist(
'../concepticondata/conceptlists/'+list2+'.tsv'
)
gidx1 = 'ENGLISH' if 'ENGLISH' in clist1['header'] else 'GLOSS'
gidx2 = 'ENGLISH' if 'ENGLISH' in clist2['header'] else 'GLOSS'
clist2_keys = []
for k in clist2:
if k not in ['header','splits','mergers']:
clist2_keys += [(clist2[k]['CONCEPTICON_ID'], k)]
clist2_keys = dict(clist2_keys)
coverage = []
count = 0
for k in clist1:
if k not in ['header', 'splits', 'mergers']:
cid = clist1[k]['CONCEPTICON_ID']
if cid in clist2_keys:
coverage += [(cid, k, clist1[k][gidx1],
clist2_keys[cid], clist2[clist2_keys[cid]][gidx2])]
count += 1
else:
coverage += [(cid, k, clist1[k][gidx1],'?','?')]
for line in coverage:
print('\t'.join(line))
print(count / len(clist1), count)
def merge_lists(*lists, output=False):
"""
Merge two or more concept lists into one single one.
"""
out = {}
for lst in lists:
clist = load_conceptlist(
'../concepticondata/conceptlists/'+lst+'.tsv'
)
gidx = 'ENGLISH' if 'ENGLISH' in clist['header'] else 'GLOSS'
for k in clist:
if k not in ['header','splits','mergers']:
gloss = clist[k][gidx]
cidx = clist[k]['CONCEPTICON_ID']
try:
out[cidx] += [gloss]
except KeyError:
out[cidx] = [gloss]
_G = nx.read_gml('../concepticondata/conceptrelations.gml')
G = nx.Graph()
for a,b,d in _G.edges(data=True):
G.add_edge(_G.node[a]['cid'], _G.node[b]['cid'], **d)
for node in out:
G.add_node(node)
# go for connected components
S = nx.subgraph(G, out)
comps = list(nx.connected_components(S))
outdata = []
C = csv2dict('../concepticondata/concepticon.tsv', strip_lines=False)
if len(comps) < len(S):
for comp in comps:
#for c in comp:
outdata += [(comp[0],out[comp[0]][0], C[comp[0]][1], ','.join([out[x][0] for x in
comp[1:]]), ','.join(C[x][1] for x in comp[1:]))]
if output == 'tsv':
with open('merger.tsv', 'w') as f:
for line in sorted(outdata, key=lambda x: x[2]):
f.write('\t'.join(line)+'\n')
else:
return outdata
def merge_lists(*lists, output=False):
"""
Merge two or more concept lists into one single one.
"""
out = {}
for lst in lists:
clist = load_conceptlist(
'../concepticondata/conceptlists/'+lst+'.tsv'
)
gidx = 'ENGLISH' if 'ENGLISH' in clist['header'] else 'GLOSS'
for k in clist:
if k not in ['header','splits','mergers']:
gloss = clist[k][gidx]
cidx = clist[k]['CONCEPTICON_ID']
try:
out[cidx] += [gloss]
except KeyError:
out[cidx] = [gloss]
_G = nx.read_gml('../concepticondata/conceptrelations.gml')
G = nx.Graph()
for a,b,d in _G.edges(data=True):
G.add_edge(_G.node[a]['cid'], _G.node[b]['cid'], **d)
for node in out:
G.add_node(node)
# go for connected components
S = nx.subgraph(G, out)
comps = list(nx.connected_components(S))
outdata = []
C = csv2dict('../concepticondata/concepticon.tsv', strip_lines=False)
if len(comps) < len(S):
for comp in comps:
#for c in comp:
outdata += [(comp[0],out[comp[0]][0], C[comp[0]][1], ','.join([out[x][0] for x in
comp[1:]]), ','.join(C[x][1] for x in comp[1:]))]
if output == 'tsv':
with open('merger.tsv', 'w') as f:
for line in sorted(outdata, key=lambda x: x[2]):
f.write('\t'.join(line)+'\n')
else:
return outdata