def test_constellation_add():
topic_query = format_cqp_query(TOPIC_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
# init constellation
topic_dump = CORPUS.query(topic_query,
context=None,
context_break=S_CONTEXT)
const = Constellation(topic_dump)
# add discourseme
disc1_query = format_cqp_query(DISC1_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
disc1_dump = CORPUS.query(disc1_query,
context=None,
context_break=S_CONTEXT)
const.add_discourseme(disc1_dump)
print(const.df)
print(const.discoursemes.keys())
def test_constellation_add(germaparl, discoursemes):
corpus = get_corpus(germaparl)
# init constellation
topic_query = format_cqp_query(
discoursemes['topic'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
topic_dump = corpus.query(
topic_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const = Constellation(topic_dump)
# add discourseme
disc1_query = format_cqp_query(
discoursemes['disc1'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
disc1_dump = corpus.query(
disc1_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const.add_discourseme(disc1_dump)
assert len(const.df) == 1599
assert 'topic' in const.discoursemes
assert len(const.discoursemes) == 2
def test_count_items(germaparl):
corpus = get_corpus(germaparl)
cqp = corpus.start_cqp()
items = ["Helmut", "Kohl", "CDU"]
queries = [format_cqp_query([item]) for item in items]
# whole corpus
counts1 = corpus.marginals(items)
counts2 = corpus.counts.mwus(cqp, queries)
assert (list(counts1["freq"]) == list(counts2["freq"]))
# subcorpus
cqp.nqr_from_query(query='[lemma="und"] expand to s', name='und')
cqp.nqr_activate(corpus.corpus_name, 'und')
counts1 = corpus.marginals(items)
counts2 = corpus.counts.mwus(cqp, queries)
assert (counts1.loc[items[0], 'freq'] > counts2.loc[queries[0], 'freq'])
# whole corpus
cqp.nqr_activate(corpus.corpus_name)
counts1 = corpus.marginals(items)
counts2 = corpus.counts.mwus(cqp, queries)
assert (list(counts1["freq"]) == list(counts2["freq"]))
cqp.__kill__()
def test_constellation_coll(germaparl, discoursemes):
corpus = get_corpus(germaparl)
# init constellation
topic_query = format_cqp_query(
discoursemes['topic'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
topic_dump = corpus.query(
topic_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const = Constellation(topic_dump)
# add discourseme 1
disc1_query = format_cqp_query(
discoursemes['disc1'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
disc1_dump = corpus.query(
disc1_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const.add_discourseme(disc1_dump, name='disc1')
# add discourseme 2
disc2_query = format_cqp_query(
discoursemes['disc2'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
disc2_dump = corpus.query(
disc2_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const.add_discourseme(disc2_dump, name='disc2')
dfs = const.collocates(windows=list(range(1, 21)))
assert len(dfs) == 20
assert len(dfs[1]) == 2
assert len(dfs[20]) == 5
def test_constellation_conc():
topic_query = format_cqp_query(TOPIC_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
# init constellation
topic_dump = CORPUS.query(topic_query,
context=None,
context_break=S_CONTEXT)
const = Constellation(topic_dump)
# add discourseme 1
disc1_query = format_cqp_query(DISC1_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
disc1_dump = CORPUS.query(disc1_query,
context=None,
context_break=S_CONTEXT)
const.add_discourseme(disc1_dump, name='disc1')
# add discourseme 2
disc2_query = format_cqp_query(DISC2_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
disc2_dump = CORPUS.query(disc2_query,
context=None,
context_break=S_CONTEXT)
const.add_discourseme(disc2_dump, name='disc2')
lines = const.concordance(s_show=['text_id'])
print(lines)
def test_count_mwus_3(germaparl):
# whole corpus
corpus = get_corpus(germaparl)
items = ["Horst Seehofer", r"( CSU )", "CSU", "WES324", "CSU"]
queries = [format_cqp_query([item]) for item in items]
cqp = corpus.start_cqp()
counts3 = corpus.counts.mwus(cqp, queries, strategy=3, fill_missing=False)
cqp.__kill__()
assert (counts3['freq']['CSU'] == 635)
def test_constellation_coll():
topic_query = format_cqp_query(TOPIC_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
# init constellation
topic_dump = CORPUS.query(topic_query,
context=None,
context_break=S_CONTEXT)
const = Constellation(topic_dump)
# add discourseme 1
disc1_query = format_cqp_query(DISC1_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
disc1_dump = CORPUS.query(disc1_query,
context=None,
context_break=S_CONTEXT)
const.add_discourseme(disc1_dump, name='disc1')
# add discourseme 2
disc2_query = format_cqp_query(DISC2_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
disc2_dump = CORPUS.query(disc2_query,
context=None,
context_break=S_CONTEXT)
const.add_discourseme(disc2_dump, name='disc2')
lines = const.collocates(windows=list(range(1, 20)))
print(lines)
def test_constellation_init():
# init constellation
topic_query = format_cqp_query(TOPIC_ITEMS,
p_query=P_QUERY,
s_query=S_QUERY,
flags="%cd",
escape=False)
topic_dump = CORPUS.query(topic_query,
context=None,
context_break=S_CONTEXT)
const = Constellation(topic_dump)
print(const.df)
def test_count_mwus_strategies(germaparl):
# whole corpus
corpus = get_corpus(germaparl)
items = ["Horst Seehofer", r"( CSU )", "CSU", "WES324"]
queries = [format_cqp_query([item]) for item in items]
cqp = corpus.start_cqp()
counts1 = corpus.counts.mwus(cqp, queries, strategy=1, fill_missing=False)
assert (queries[0] in counts1.index)
counts2 = corpus.counts.mwus(cqp, queries, strategy=2, fill_missing=False)
counts3 = corpus.counts.mwus(cqp, queries, strategy=3, fill_missing=False)
cqp.__kill__()
assert (counts2.equals(counts3))
assert (sum(counts1['freq']) == sum(counts2['freq']))
def test_count_items_subcorpora(germaparl):
# subcorpus
corpus = get_corpus(germaparl)
cqp = corpus.start_cqp()
dump = corpus.dump_from_s_att("text_role", ["presidency"])
cqp.nqr_from_dump(dump, 'Presidency')
cqp.nqr_activate(corpus.corpus_name, 'Presidency')
items = ["Horst Seehofer", r"( CSU )", "CSU", "WES324", "CSU"]
queries = [format_cqp_query([item]) for item in items]
counts1 = corpus.counts.mwus(cqp, queries, strategy=1, fill_missing=False)
assert (sum(counts1['freq']) > 0)
counts2 = corpus.counts.mwus(cqp, queries, strategy=2, fill_missing=False)
counts3 = corpus.counts.mwus(cqp, queries, strategy=3, fill_missing=False)
assert (counts2.equals(counts3))
cqp.__kill__()
def test_constellation_init(germaparl, discoursemes):
corpus = get_corpus(germaparl)
# init constellation
topic_query = format_cqp_query(
discoursemes['topic'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
topic_dump = corpus.query(
topic_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const = Constellation(topic_dump)
assert isinstance(const.df, DataFrame)
assert len(const.df) == 2777
def test_textual_constellation(germaparl, discoursemes):
corpus = get_corpus(germaparl)
# init constellation
topic_query = format_cqp_query(
discoursemes['topic'],
p_query=discoursemes['parameters']['p_query'],
s_query=discoursemes['parameters']['s_query'],
flags=discoursemes['parameters']['flags_query'],
escape=discoursemes['parameters']['escape_query'])
topic_dump = corpus.query(
topic_query,
context=None,
context_break=discoursemes['parameters']['s_context'])
const = TextConstellation(
topic_dump, s_context=discoursemes['parameters']['s_context'])
assert len(const.df) == 624
assert 'MATCHES_topic' in const.df.columns