def test_frequency_analysis_fast0(self):
with timer() as t1:
rows1 = self.query('''
SELECT fn1.performance_reduce_characters_fast(w, c)
FROM (
SELECT fn1.performance_map_characters_fast0(text)
FROM daten.wiki_names
WHERE text IS NOT NULL
)
GROUP BY w
ORDER BY c DESC, w ASC''')
with timer() as t2:
rows2 = self.query('''
SELECT w, c
FROM daten.wiki_freq
ORDER BY c DESC, w ASC''')
data = [tuple(x) for x in rows1]
reference = [tuple(x) for x in rows2]
print "test_frequency_analysis_fast0 query:", t1.duration, t2.duration
self.compare(reference, data)
def test_frequency_analysis(self):
if udf.opts.lang == 'r':
raise SkipTest('this R implementation is too slow')
with timer() as t1:
rows1 = self.query('''
SELECT fn1.performance_reduce_characters(w, c)
FROM (
SELECT fn1.performance_map_characters(text)
FROM daten.wiki_names
)
GROUP BY w
ORDER BY c DESC, w ASC''')
with timer() as t2:
rows2 = self.query('''
SELECT w, c
FROM daten.wiki_freq
ORDER BY c DESC, w ASC''')
data = [tuple(x) for x in rows1]
reference = [tuple(x) for x in rows2]
print "test_frequency_analysis query:", t1.duration, t2.duration
self.compare(reference, data)
def test_word_unicode_count(self):
'''DWA-13860 (lua)'''
sql = '''
SELECT performance_reduce_counts(w, c)
FROM (
SELECT performance_map_unicode_words(c3_varchar100)
FROM test.enginetablebigunicode
)
GROUP BY w
ORDER BY 1 DESC'''
with timer() as t:
self.query(sql)
self.assertLessEqual(t.duration, 11)
def test_word_count_fast77(self):
sql = '''
SELECT COUNT(*) FROM (
SELECT performance_reduce_counts_fast77(w, c)
FROM (
SELECT performance_map_words(varchar02)
FROM test.enginetablebig1
)
GROUP BY w
ORDER BY 1 DESC)'''
with timer() as t:
ret = self.query(sql)
print "test_word_count_fast77 query:", t.duration, repr(ret)
self.assertLessEqual(t.duration, 160)