def data_frame_group_by_aggregate_14668():
data = data_frame("vehicles.csv")
start = time.time()
(data
.group_by("make", "model", "year")
.aggregate(
n=di.count(),
hwy=di.mean("hwy"),
cty=di.mean("cty")))
return time.time() - start
def data_frame_group_by_aggregate_100000_short_numba():
with patch("dataiter.USE_NUMBA", True):
data = data_frame_random(1_000_000, 100_000)
start = time.time()
(data
.group_by("g")
.aggregate(
a_mean=di.mean("a"),
a_std=di.std("a")))
return time.time() - start
def test_aggregate(self):
data = test.data_frame("vehicles.csv")
data = data.filter_out(data.cyl.is_na())
data = data.filter_out(data.displ.is_na())
stat = (data.group_by("make",
"model").aggregate(cyl=di.median("cyl"),
displ=di.mean("displ")))
assert stat.nrow == 3240
assert stat.ncol == 4
assert stat.sort(make=1, model=1) == stat
assert np.isclose(np.sum(stat.cyl), 19964.5, atol=0.1)
assert np.isclose(np.sum(stat.displ), 11430.1, atol=0.1)
data[name] = np.char.lower(data[name])
return data
# AGGREGATE
(read_csv("../data/vehicles.csv")
.modify(fuel_regular=lambda x: x.fuel == "regular")
.group_by("make", "model")
.aggregate(
all_fuel_regular=di.all("fuel_regular"),
any_fuel_regular=di.any("fuel_regular"),
count=di.count(),
count_unique_cyl=di.count_unique("cyl"),
first_hwy=di.first("hwy"),
last_hwy=di.last("hwy"),
max_hwy=di.max("hwy"),
mean_hwy=di.mean("hwy"),
median_hwy=di.median("hwy"),
min_hwy=di.min("hwy"),
mode_year=di.mode("year"),
nth_id=di.nth("id", 0),
quantile_hwy=di.quantile("hwy", 0.75),
std_hwy=di.std("hwy", ddof=1),
sum_hwy=di.sum("hwy"),
var_hwy=di.var("hwy", ddof=1))
.modify(mean_hwy=lambda x: x.mean_hwy.round(2))
.modify(std_hwy =lambda x: x.std_hwy.round(2))
.modify(var_hwy =lambda x: x.var_hwy.round(2))
.write_csv("aggregate.df.csv"))
# ANTI JOIN
reviews = read_csv("../data/listings-reviews.csv")