covarSampStable
Introduit dans : v1.1.0
Calcule la covariance d’échantillon :
\frac{\Sigma{(x - \bar{x})(y - \bar{y})}}{n - 1}Cette fonction est similaire à covarSamp, mais utilise un algorithme numériquement stable.
En conséquence, covarSampStable est plus lente que covarSamp, mais produit une erreur de calcul plus faible.
Syntaxe
covarSampStable(x, y)Arguments
x— Première variable.(U)Int*ouFloat*ouDecimaly— Deuxième variable.(U)Int*ouFloat*ouDecimal
Valeur renvoyée
Renvoie la covariance d’échantillon entre x et y. Pour n <= 1, inf est renvoyé. Float64
Exemples
Calcul de base de la covariance d’échantillon avec un algorithme numériquement stable
DROP TABLE IF EXISTS series;
CREATE TABLE series(i UInt32, x_value Float64, y_value Float64) ENGINE = Memory;
INSERT INTO series(i, x_value, y_value) VALUES (1, 5.6,-4.4),(2, -9.6,3),(3, -1.3,-4),(4, 5.3,9.7),(5, 4.4,0.037),(6, -8.6,-7.8),(7, 5.1,9.3),(8, 7.9,-3.6),(9, -8.2,0.62),(10, -3,7.3);
SELECT covarSampStable(x_value, y_value)
FROM
(
SELECT
x_value,
y_value
FROM series
);┌─covarSampStable(x_value, y_value)─┐
│ 7.206275555555556 │
└───────────────────────────────────┘Une valeur unique renvoie inf
SELECT covarSampStable(x_value, y_value)
FROM
(
SELECT
x_value,
y_value
FROM series LIMIT 1
);┌─covarSampStable(x_value, y_value)─┐
│ inf │
└───────────────────────────────────┘