# scatterplot
sns.relplot(data=penguins,
x='flipper_length_mm', y='body_mass_g',
hue='species', style='sex')
from sklearn.linear_model import LinearRegression
x = data.x.values.reshape(-1,1)
y = data['y']
model = LinearRegression().fit(x, y)
print(f"Regression line: ŷ = {model.intercept_:.2f} + {model.coef_[0]:.2f} x")
# Predict y values corresponding to x
model.predict([[valueOpX]])[0]
correlation coefficient and the coefficient of determination.
Explained variance | Linear relation | ||
---|---|---|---|
< .3 | < .1 | < 10% | very weak |
.3 - .5 | .1 - .25 | 10% - 25% | weak |
.5 - .7 | .25 - .5 | 25% - 50% | moderate |
.7 - .85 | .5 - .75 | 50% - 75% | strong |
.85 - .95 | .75 - .9 | 75% - 90% | very strong |
> .95 | > .9 | > 90% | exceptionally strong |
cor = np.corrcoef(cats.Hwt, cats.Bwt)[0][1]
print(f"R = { cor }")
print(f"R² = {cor ** 2}")