thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-23 08:02:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

update test

Browse Source
This commit is contained in:
nikitamikhaylov 2020-11-25 17:45:27 +03:00
parent 02a80230f8
commit 44f4064e9b
3 changed files with 14 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/AggregateFunctions/AggregateFunctionStudentTTest.cpp
View File

@ -47,7 +47,7 @@ struct StudentTTestData : public TTestMoments<Float64>
Float64 all_y = y2 + ny * mean_y * mean_y - 2 * mean_y * y1;
Float64 s2 = (all_x + all_y) / degrees_of_freedom;
Float64 std_err2 = s2 * (1 / nx + 1 / ny);
Float64 std_err2 = s2 * (1. / nx + 1. / ny);
/// t-statistic
Float64 t_stat = (mean_x - mean_y) / sqrt(std_err2);

2
src/AggregateFunctions/StatCommon.h
View File

@ -14,7 +14,7 @@ namespace DB
template <typename F>
static Float64 integrateSimpson(Float64 a, Float64 b, F && func)
{
const size_t iterations = std::max(1e6, 1e4 * std::abs(std::round(b)));
const size_t iterations = std::max(1e6, 1e4 * std::abs(std::round(b) - std::round(a)));
const long double h = (b - a) / iterations;
Float64 sum_odds = 0.0;
for (size_t i = 1; i < iterations; i += 2)

24
tests/queries/0_stateless/01558_ttest_scipy.python
View File

@ -10,19 +10,19 @@ sys.path.insert(0, os.path.join(CURDIR, 'helpers'))
from pure_http_client import ClickHouseClient
def test_and_check(name, a, b, t_stat, p_value, precision=5 * 1e-2):
def test_and_check(name, a, b, t_stat, p_value, precision=1e-2):
client = ClickHouseClient()
client.query("DROP TABLE IF EXISTS ttest;")
client.query("CREATE TABLE ttest (left Float64, right UInt8) ENGINE = Memory;");
insert_query = "INSERT INTO ttest VALUES {};".format(", ".join(['({},{}), ({},{})'.format(i, 0, j, 1) for i,j in zip(a, b)]))
client.query(insert_query)
client.query("INSERT INTO ttest VALUES {};".format(", ".join(['({},{})'.format(i, 0) for i in a])))
client.query("INSERT INTO ttest VALUES {};".format(", ".join(['({},{})'.format(j, 1) for j in b])))
real = client.query_return_df(
"SELECT roundBankers({}(left, right).1, 16) as t_stat, ".format(name) +
"roundBankers({}(left, right).2, 16) as p_value ".format(name) +
"FROM ttest FORMAT TabSeparatedWithNames;")
real_t_stat = real['t_stat'][0]
real_p_value = real['p_value'][0]
assert(abs(real_t_stat - np.float64(t_stat) < precision)), "clickhouse_t_stat {}, scipy_t_stat {}".format(real_t_stat, t_stat)
assert(abs(real_t_stat - np.float64(t_stat)) < precision), "clickhouse_t_stat {}, scipy_t_stat {}".format(real_t_stat, t_stat)
assert(abs(real_p_value - np.float64(p_value)) < precision), "clickhouse_p_value {}, scipy_p_value {}".format(real_p_value, p_value)
client.query("DROP TABLE IF EXISTS ttest;")
@ -38,13 +38,13 @@ def test_student():
s, p = stats.ttest_ind(rvs1, rvs2, equal_var = True)
test_and_check("studentTTest", rvs1, rvs2, s, p)
rvs1 = np.round(stats.norm.rvs(loc=2, scale=10,size=65536), 2)
rvs2 = np.round(stats.norm.rvs(loc=5, scale=20,size=65536), 2)
rvs1 = np.round(stats.norm.rvs(loc=2, scale=10,size=512), 2)
rvs2 = np.round(stats.norm.rvs(loc=5, scale=20,size=1024), 2)
s, p = stats.ttest_ind(rvs1, rvs2, equal_var = True)
test_and_check("studentTTest", rvs1, rvs2, s, p)
rvs1 = np.round(stats.norm.rvs(loc=0, scale=10,size=65536), 2)
rvs2 = np.round(stats.norm.rvs(loc=0, scale=10,size=65536), 2)
rvs1 = np.round(stats.norm.rvs(loc=0, scale=10,size=1024), 2)
rvs2 = np.round(stats.norm.rvs(loc=0, scale=10,size=512), 2)
s, p = stats.ttest_ind(rvs1, rvs2, equal_var = True)
test_and_check("studentTTest", rvs1, rvs2, s, p)
@ -59,13 +59,13 @@ def test_welch():
s, p = stats.ttest_ind(rvs1, rvs2, equal_var = False)
test_and_check("welchTTest", rvs1, rvs2, s, p)
rvs1 = np.round(stats.norm.rvs(loc=0, scale=10,size=65536), 2)
rvs2 = np.round(stats.norm.rvs(loc=5, scale=1,size=65536), 2)
rvs1 = np.round(stats.norm.rvs(loc=0, scale=10,size=1024), 2)
rvs2 = np.round(stats.norm.rvs(loc=5, scale=1,size=512), 2)
s, p = stats.ttest_ind(rvs1, rvs2, equal_var = False)
test_and_check("welchTTest", rvs1, rvs2, s, p)
rvs1 = np.round(stats.norm.rvs(loc=5, scale=10,size=65536), 2)
rvs2 = np.round(stats.norm.rvs(loc=5, scale=10,size=65536), 2)
rvs1 = np.round(stats.norm.rvs(loc=5, scale=10,size=512), 2)
rvs2 = np.round(stats.norm.rvs(loc=5, scale=10,size=1024), 2)
s, p = stats.ttest_ind(rvs1, rvs2, equal_var = False)
test_and_check("welchTTest", rvs1, rvs2, s, p)