def demographics(con):
data = {}
cur = con.cursor()
cur.execute("SELECT gender, birth_year, COUNT(trip_id) trip_count "
"FROM trips WHERE gender NOT NULL AND birth_year NOT NULL "
"GROUP BY gender, birth_year ORDER BY trip_count DESC")
while True:
rows = cur.fetchmany(1000)
if not rows: break
for row in rows:
put_val(data, row["gender"], row["birth_year"], row["trip_count"])
cur.close()
for kvs, data in traverse(data, "gender"):
gender = kvs["gender"]
xs = list(sorted(data.keys()))
ys = [data[x] for x in xs]
total = sum(ys)
fig = plt.figure()
plt.scatter(xs, ys)
plt.yscale("log")
plt.xlabel("birth year")
plt.ylabel("number of rides")
plt.grid(axis="y")
plt.title("%s n=%d" % (gender, total))
plt.savefig("gender.%s.pdf" % gender, bbox_inches="tight")
def demographics(con):
data = {}
cur = con.cursor()
cur.execute("SELECT gender, birth_year, COUNT(trip_id) trip_count "
"FROM trips WHERE gender NOT NULL AND birth_year NOT NULL "
"GROUP BY gender, birth_year ORDER BY trip_count DESC")
while True:
rows = cur.fetchmany(1000)
if not rows: break
for row in rows:
put_val(data, row["gender"], row["birth_year"], row["trip_count"])
cur.close()
for kvs, data in traverse(data, "gender"):
gender = kvs["gender"]
xs = list(sorted(data.keys()))
ys = [data[x] for x in xs]
total = sum(ys)
fig = plt.figure()
plt.scatter(xs, ys)
plt.yscale("log")
plt.xlabel("birth year")
plt.ylabel("number of rides")
plt.grid(axis="y")
plt.title("%s n=%d" % (gender, total))
plt.savefig("gender.%s.pdf" % gender, bbox_inches="tight")
def collect_nested_loops(dostmt):
bag = odict({"top": dostmt.parent, "dopair": {dostmt.parent: dostmt}})
ret = traverse(dostmt.parent, search_nested_loops, bag, subnode="content", prerun=True)
nested = odict()
queue = [(dostmt.parent, nested)]
while queue:
parent, nestdict = queue.pop(0)
nesteddict = odict()
nestdict[bag["dopair"][parent]] = nesteddict
if parent in bag:
queue.extend([(p, nesteddict) for p in bag[parent]])
return nested
def write_test_set(self,f,test_set):
traverse(lambda x: util.write_line(f,str(' '.join(nltk.word_tokenize(x)))),test_set)